Popular Wordle Openers.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "b33f69da",
   "metadata": {},
   "source": [
    "## Wordle Opener Analysis from Tweets\n",
    "\n",
    "Wordle tweet data is a useful data source and can, for example, [accurately predict](https://twitter.com/thewordlebot/status/1498661373626265600) the wordle solution for the day. \n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fdf684d3",
   "metadata": {},
   "source": [
    "\n",
    "Analyzing the popularity of wordle openers is a little harder, since on any given day all we know are the most popular opening patterns.) I used a Ridge linear regression to look for popular openers across multiple wordles.\n",
    "\n",
    "TDLR; the most common openers are below, though bear in mind even the top choice `adieu` is estimated to be only 1.2-5% of all openers. \n",
    "\n",
    "1. adieu\n",
    "2. audio\n",
    "3. stare\n",
    "4. crane\n",
    "5. arise\n",
    "6. irate\n",
    "7. heart\n",
    "8. addio [^1]\n",
    "9. raise\n",
    "10. great\n",
    "11. train\n",
    "12. arose\n",
    "13. crate\n",
    "14. tears\n",
    "15. soare\n",
    "[^1]: A false positive due to similarity with audio.\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eef90487",
   "metadata": {},
   "source": [
    "\n",
    "\n",
    "### Load and prep the data\n",
    "Most of the functionality used is in the files `first_word.py`. The function `make_first_guess_list` creates a useful dataframe for analysis.\n",
    "\n",
    "I start with the excellent kaggle data set [wordle-tweets](https://www.kaggle.com/benhamner/wordle-tweets) and extract directly from the zipfile which I download with the [kaggle api](https://github.com/Kaggle/kaggle-api). The `get_first_words` function processes the dataframe:\n",
    "\n",
    "    1. It removes tweeted scores that contain an invalid score line for that answer (likely played on a cached version of the puzzle, not the live one on the NY Times site).\n",
    "    2. It removes any tweets that have more than 6 score patterns.\n",
    "    3. It extracts the first pattern from the tweeted scores.\n",
    "    4. It maps on the answer for a given wordle id.\n",
    "    5. It groups by the answer for the day, and creates a dataframe that has score, target, guess, and some data on how popular that score is."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "26b3af68",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:40:22.941562Z",
     "start_time": "2022-03-22T00:40:12.917777Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Max wordle num 294\n",
      "Filtered out 19057 of 424360 rows\n"
     ]
    }
   ],
   "source": [
    "from first_word import make_first_guest_list\n",
    "df = make_first_guest_list()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "2147399f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:23:32.550718Z",
     "start_time": "2022-03-22T00:23:32.530900Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>score</th>\n",
       "      <th>target</th>\n",
       "      <th>guess</th>\n",
       "      <th>score_frequency_rank</th>\n",
       "      <th>score_count_fraction</th>\n",
       "      <th>wordle_num</th>\n",
       "      <th>guess_count</th>\n",
       "      <th>commonality</th>\n",
       "      <th>weighted_rank</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>21000</td>\n",
       "      <td>ahead</td>\n",
       "      <td>aalii</td>\n",
       "      <td>60.0</td>\n",
       "      <td>0.001005</td>\n",
       "      <td>258</td>\n",
       "      <td>23</td>\n",
       "      <td>0</td>\n",
       "      <td>3600.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>21001</td>\n",
       "      <td>ahead</td>\n",
       "      <td>aargh</td>\n",
       "      <td>52.0</td>\n",
       "      <td>0.001724</td>\n",
       "      <td>258</td>\n",
       "      <td>7</td>\n",
       "      <td>71592</td>\n",
       "      <td>2704.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>21000</td>\n",
       "      <td>ahead</td>\n",
       "      <td>aarti</td>\n",
       "      <td>60.0</td>\n",
       "      <td>0.001005</td>\n",
       "      <td>258</td>\n",
       "      <td>23</td>\n",
       "      <td>63273</td>\n",
       "      <td>3600.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>20100</td>\n",
       "      <td>ahead</td>\n",
       "      <td>abaca</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.040937</td>\n",
       "      <td>258</td>\n",
       "      <td>63</td>\n",
       "      <td>79841</td>\n",
       "      <td>100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>20100</td>\n",
       "      <td>ahead</td>\n",
       "      <td>abaci</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.040937</td>\n",
       "      <td>258</td>\n",
       "      <td>63</td>\n",
       "      <td>0</td>\n",
       "      <td>100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12940</th>\n",
       "      <td>01000</td>\n",
       "      <td>wrung</td>\n",
       "      <td>zuzim</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.050536</td>\n",
       "      <td>225</td>\n",
       "      <td>852</td>\n",
       "      <td>0</td>\n",
       "      <td>49.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12941</th>\n",
       "      <td>00100</td>\n",
       "      <td>wrung</td>\n",
       "      <td>zygal</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.069694</td>\n",
       "      <td>225</td>\n",
       "      <td>1461</td>\n",
       "      <td>0</td>\n",
       "      <td>25.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12942</th>\n",
       "      <td>00101</td>\n",
       "      <td>wrung</td>\n",
       "      <td>zygon</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.017046</td>\n",
       "      <td>225</td>\n",
       "      <td>153</td>\n",
       "      <td>15300</td>\n",
       "      <td>121.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12943</th>\n",
       "      <td>00000</td>\n",
       "      <td>wrung</td>\n",
       "      <td>zymes</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.173631</td>\n",
       "      <td>225</td>\n",
       "      <td>4320</td>\n",
       "      <td>24137</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12944</th>\n",
       "      <td>00000</td>\n",
       "      <td>wrung</td>\n",
       "      <td>zymic</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.173631</td>\n",
       "      <td>225</td>\n",
       "      <td>4320</td>\n",
       "      <td>30459</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1074435 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       score target  guess  score_frequency_rank  score_count_fraction  \\\n",
       "0      21000  ahead  aalii                  60.0              0.001005   \n",
       "1      21001  ahead  aargh                  52.0              0.001724   \n",
       "2      21000  ahead  aarti                  60.0              0.001005   \n",
       "3      20100  ahead  abaca                  10.0              0.040937   \n",
       "4      20100  ahead  abaci                  10.0              0.040937   \n",
       "...      ...    ...    ...                   ...                   ...   \n",
       "12940  01000  wrung  zuzim                   7.0              0.050536   \n",
       "12941  00100  wrung  zygal                   5.0              0.069694   \n",
       "12942  00101  wrung  zygon                  11.0              0.017046   \n",
       "12943  00000  wrung  zymes                   1.0              0.173631   \n",
       "12944  00000  wrung  zymic                   1.0              0.173631   \n",
       "\n",
       "       wordle_num  guess_count  commonality  weighted_rank  \n",
       "0             258           23            0         3600.0  \n",
       "1             258            7        71592         2704.0  \n",
       "2             258           23        63273         3600.0  \n",
       "3             258           63        79841          100.0  \n",
       "4             258           63            0          100.0  \n",
       "...           ...          ...          ...            ...  \n",
       "12940         225          852            0           49.0  \n",
       "12941         225         1461            0           25.0  \n",
       "12942         225          153        15300          121.0  \n",
       "12943         225         4320        24137            1.0  \n",
       "12944         225         4320        30459            1.0  \n",
       "\n",
       "[1074435 rows x 9 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df\n",
    "import plotly.io as pio\n",
    "pio.renderers.default = \"png\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "196a6257",
   "metadata": {},
   "source": [
    "### Simple analysis\n",
    "\n",
    "We have a large dataframe of almost 600,000 columns. One row for every pattern and every guess for every wordle answer. A simple way to look at common starter words would be to group by the guess and look which words consistently rank high across several days."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "b033a4da",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:23:32.596409Z",
     "start_time": "2022-03-22T00:23:32.552125Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "guess\n",
       "adieu    2.481928\n",
       "braze    3.349398\n",
       "graze    3.433735\n",
       "craze    3.542169\n",
       "brave    4.566265\n",
       "grave    4.650602\n",
       "crave    4.759036\n",
       "brane    4.908537\n",
       "audio    4.933735\n",
       "crare    4.957317\n",
       "Name: score_frequency_rank, dtype: float64"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby('guess')['score_frequency_rank'].mean().sort_values().head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d0c269e5",
   "metadata": {},
   "source": [
    "This analysis leaves a lot to be desired. While other evidence indicates `adieu` is a popular opener, it does not seem like `jazzy` would be one. Or for that matter, words with a `z`. \n",
    "\n",
    "The `00000` all grey pattern is fairly common, so words with uncommon letters will show up a lot since `00000` is common because of the sheer volume of words that can create the pattern. So what happens if we  filter out the null score? Not much..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "5b973b6b",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:23:32.705694Z",
     "start_time": "2022-03-22T00:23:32.597191Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "guess\n",
       "adieu    2.662162\n",
       "braze    4.046875\n",
       "craze    4.102941\n",
       "graze    4.107692\n",
       "audio    5.173077\n",
       "brane    5.514085\n",
       "crave    5.521739\n",
       "brave    5.553846\n",
       "grave    5.590909\n",
       "grate    5.725352\n",
       "Name: score_frequency_rank, dtype: float64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.query(\"score != '00000'\").groupby('guess')['score_frequency_rank'].mean().sort_values().head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fad6413a",
   "metadata": {},
   "source": [
    "So are people guessing `braze` first regularly or is something else going on? BRAZE's score line is common but what other words could make the same patern?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "d34d20de",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:23:32.744479Z",
     "start_time": "2022-03-22T00:23:32.706620Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>score</th>\n",
       "      <th>target</th>\n",
       "      <th>guess</th>\n",
       "      <th>score_frequency_rank</th>\n",
       "      <th>score_count_fraction</th>\n",
       "      <th>wordle_num</th>\n",
       "      <th>guess_count</th>\n",
       "      <th>commonality</th>\n",
       "      <th>weighted_rank</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5315</th>\n",
       "      <td>00101</td>\n",
       "      <td>pleat</td>\n",
       "      <td>image</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.029706</td>\n",
       "      <td>230</td>\n",
       "      <td>186</td>\n",
       "      <td>197874283</td>\n",
       "      <td>81.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9847</th>\n",
       "      <td>00101</td>\n",
       "      <td>pleat</td>\n",
       "      <td>share</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.029706</td>\n",
       "      <td>230</td>\n",
       "      <td>186</td>\n",
       "      <td>119294241</td>\n",
       "      <td>81.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11916</th>\n",
       "      <td>00101</td>\n",
       "      <td>pleat</td>\n",
       "      <td>until</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.029706</td>\n",
       "      <td>230</td>\n",
       "      <td>186</td>\n",
       "      <td>113090086</td>\n",
       "      <td>81.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4576</th>\n",
       "      <td>00101</td>\n",
       "      <td>pleat</td>\n",
       "      <td>grade</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.029706</td>\n",
       "      <td>230</td>\n",
       "      <td>186</td>\n",
       "      <td>54275130</td>\n",
       "      <td>81.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4011</th>\n",
       "      <td>00101</td>\n",
       "      <td>pleat</td>\n",
       "      <td>frame</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.029706</td>\n",
       "      <td>230</td>\n",
       "      <td>186</td>\n",
       "      <td>46079991</td>\n",
       "      <td>81.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11974</th>\n",
       "      <td>00101</td>\n",
       "      <td>pleat</td>\n",
       "      <td>usage</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.029706</td>\n",
       "      <td>230</td>\n",
       "      <td>186</td>\n",
       "      <td>25440406</td>\n",
       "      <td>81.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9850</th>\n",
       "      <td>00101</td>\n",
       "      <td>pleat</td>\n",
       "      <td>sharp</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.029706</td>\n",
       "      <td>230</td>\n",
       "      <td>186</td>\n",
       "      <td>24904199</td>\n",
       "      <td>81.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4575</th>\n",
       "      <td>00101</td>\n",
       "      <td>pleat</td>\n",
       "      <td>grace</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.029706</td>\n",
       "      <td>230</td>\n",
       "      <td>186</td>\n",
       "      <td>17642126</td>\n",
       "      <td>81.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12119</th>\n",
       "      <td>00101</td>\n",
       "      <td>pleat</td>\n",
       "      <td>villa</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.029706</td>\n",
       "      <td>230</td>\n",
       "      <td>186</td>\n",
       "      <td>17587586</td>\n",
       "      <td>81.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10389</th>\n",
       "      <td>00101</td>\n",
       "      <td>pleat</td>\n",
       "      <td>solve</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.029706</td>\n",
       "      <td>230</td>\n",
       "      <td>186</td>\n",
       "      <td>13452150</td>\n",
       "      <td>81.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       score target  guess  score_frequency_rank  score_count_fraction  \\\n",
       "5315   00101  pleat  image                   9.0              0.029706   \n",
       "9847   00101  pleat  share                   9.0              0.029706   \n",
       "11916  00101  pleat  until                   9.0              0.029706   \n",
       "4576   00101  pleat  grade                   9.0              0.029706   \n",
       "4011   00101  pleat  frame                   9.0              0.029706   \n",
       "11974  00101  pleat  usage                   9.0              0.029706   \n",
       "9850   00101  pleat  sharp                   9.0              0.029706   \n",
       "4575   00101  pleat  grace                   9.0              0.029706   \n",
       "12119  00101  pleat  villa                   9.0              0.029706   \n",
       "10389  00101  pleat  solve                   9.0              0.029706   \n",
       "\n",
       "       wordle_num  guess_count  commonality  weighted_rank  \n",
       "5315          230          186    197874283           81.0  \n",
       "9847          230          186    119294241           81.0  \n",
       "11916         230          186    113090086           81.0  \n",
       "4576          230          186     54275130           81.0  \n",
       "4011          230          186     46079991           81.0  \n",
       "11974         230          186     25440406           81.0  \n",
       "9850          230          186     24904199           81.0  \n",
       "4575          230          186     17642126           81.0  \n",
       "12119         230          186     17587586           81.0  \n",
       "10389         230          186     13452150           81.0  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.query('score == \"00101\" and wordle_num == 230').sort_values('commonality',ascending=False).head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "31c06d7d",
   "metadata": {},
   "source": [
    "There are a lot of words that make the same pattern, is `BRAZE` popular or is it riding the coattails of some other `_RA_E` word?"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0fa42b53",
   "metadata": {},
   "source": [
    "### Linear Regression\n",
    "\n",
    "A better approach is to control for the presence of other words. If BRAZE only does well when it's paired with GRACE or GRADE or SHARE than a linear regression should isolate the guesses that actually are predictive of a popular score count line. But, there will be a lot of colinearity so we will use a Ridge regression.\n",
    "\n",
    "The difficulty comes from getting the dataframe into the right format, thankfully I discovered [pd.crosstab](https://pandas.pydata.org/docs/reference/api/pandas.crosstab.html), and there are [other methods that were all better](https://stackoverflow.com/questions/46791626/one-hot-encoding-multi-level-column-data) than what I had been doing originally (an awful groupby loop than took over a minute.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "539c0da3",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:31:20.165962Z",
     "start_time": "2022-03-22T00:31:09.875687Z"
    }
   },
   "outputs": [],
   "source": [
    "df.rename(columns={'score': 'score_pattern'}, inplace=True)\n",
    "\n",
    "one_hot_encoded_data = pd.crosstab(\n",
    "    [df['wordle_num'], df['score_pattern']], df['guess']).join(\n",
    "        df.groupby(['wordle_num',\n",
    "                    'score_pattern'])[['score_count_fraction',\n",
    "                                       'guess_count']].first()).reset_index()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7b1ca61a",
   "metadata": {},
   "source": [
    "Now we'll sort of normalize the data a little and then fit a model. I will use a [Ridge](https://scikit-learn.org/stable/modules/linear_model.html#ridge-regression-and-classification) which helps control the size of the coefficients, and is neccessary to handle the colinearity of the variables."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "84e6b296",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:31:20.699551Z",
     "start_time": "2022-03-22T00:31:20.167134Z"
    }
   },
   "outputs": [],
   "source": [
    "one_hot_encoded_data.dropna(subset=['score_count_fraction'],inplace=True) #don't need patterns no one actually guessed\n",
    "# actually not sure if fillna 0 would be better?\n",
    "std = one_hot_encoded_data['score_count_fraction'].std()\n",
    "# mean = one_hot_encoded_data['score_count_fraction'].mean()\n",
    "# one_hot_encoded_data['score_count_fraction'] = (one_hot_encoded_data[\n",
    "#     'score_count_fraction'] - mean) / std\n",
    "\n",
    "guess_count_std = one_hot_encoded_data[\n",
    "    'guess_count'].std()\n",
    "guess_count_mean = one_hot_encoded_data['guess_count'].mean()\n",
    "\n",
    "one_hot_encoded_data['guess_count'] = (one_hot_encoded_data[\n",
    "    'guess_count'] - guess_count_mean ) / guess_count_std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "5acb1200",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:31:20.719501Z",
     "start_time": "2022-03-22T00:31:20.700308Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        0.099927\n",
       "1        0.015430\n",
       "3        0.092212\n",
       "4        0.002939\n",
       "6        0.008082\n",
       "           ...   \n",
       "11831    0.001766\n",
       "11833    0.007357\n",
       "11834    0.000294\n",
       "11835    0.032961\n",
       "11836    0.002649\n",
       "Name: score_count_fraction, Length: 8061, dtype: float64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "one_hot_encoded_data['score_count_fraction']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "643f3fb8",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:31:52.847885Z",
     "start_time": "2022-03-22T00:31:23.802087Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RidgeCV(alphas=array([10, 25, 35, 50]))"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import linear_model\n",
    "data = one_hot_encoded_data.query(\"wordle_num != 258\") #explained further down, this data point has particularly bad colinearity issues\n",
    "X= data.drop(\n",
    "        columns=[ 'score_count_fraction', 'wordle_num','score_pattern']) # fit to the one hot encoded guesses and the total guess count\n",
    "y=data['score_count_fraction'] # our dependent is the fraction of guesses that had the score pattern\n",
    "\n",
    "r = linear_model.RidgeCV(alphas=[10,25,35,50])\n",
    "r.fit(X,y)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c38a873",
   "metadata": {},
   "source": [
    "#### Results!\n",
    "\n",
    "Now we can just look at the coefficients of the variables to see which ones most strongly predict a popular pattern. The top few guesses contain [some](https://yougov.co.uk/topics/entertainment/articles-reports/2022/02/03/wordle-starter-words-hard-mode-and-x6-how-are-brit) [familiar](https://today.yougov.com/topics/lifestyle/articles-reports/2022/02/18/wordle-willingness-to-pay-poll) choices. \n",
    "\n",
    "Though to repeat, `addio` is almost certainly not real but riding the coattails of `audio`.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "3e027634",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:31:52.872010Z",
     "start_time": "2022-03-22T00:31:52.849440Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r.alpha_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "c98a9b58",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:31:52.904358Z",
     "start_time": "2022-03-22T00:31:52.873342Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>coef</th>\n",
       "      <th>variable</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>0.017741</td>\n",
       "      <td>adieu</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>633</th>\n",
       "      <td>0.011133</td>\n",
       "      <td>audio</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10645</th>\n",
       "      <td>0.008327</td>\n",
       "      <td>stare</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2362</th>\n",
       "      <td>0.007077</td>\n",
       "      <td>crane</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>531</th>\n",
       "      <td>0.006380</td>\n",
       "      <td>arise</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5401</th>\n",
       "      <td>0.006203</td>\n",
       "      <td>irate</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4926</th>\n",
       "      <td>0.005777</td>\n",
       "      <td>heart</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>103</th>\n",
       "      <td>0.005218</td>\n",
       "      <td>addio</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8839</th>\n",
       "      <td>0.005135</td>\n",
       "      <td>raise</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4602</th>\n",
       "      <td>0.005063</td>\n",
       "      <td>great</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11564</th>\n",
       "      <td>0.004999</td>\n",
       "      <td>train</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>547</th>\n",
       "      <td>0.004940</td>\n",
       "      <td>arose</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2371</th>\n",
       "      <td>0.004466</td>\n",
       "      <td>crate</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11158</th>\n",
       "      <td>0.004272</td>\n",
       "      <td>tears</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10342</th>\n",
       "      <td>0.004050</td>\n",
       "      <td>soare</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           coef variable\n",
       "108    0.017741    adieu\n",
       "633    0.011133    audio\n",
       "10645  0.008327    stare\n",
       "2362   0.007077    crane\n",
       "531    0.006380    arise\n",
       "5401   0.006203    irate\n",
       "4926   0.005777    heart\n",
       "103    0.005218    addio\n",
       "8839   0.005135    raise\n",
       "4602   0.005063    great\n",
       "11564  0.004999    train\n",
       "547    0.004940    arose\n",
       "2371   0.004466    crate\n",
       "11158  0.004272    tears\n",
       "10342  0.004050    soare"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_openers = pd.DataFrame(list(zip(r.coef_, r.feature_names_in_)),\n",
    "         columns=['coef', 'variable']).sort_values('coef',ascending=False)\n",
    "top_openers.head(15)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8bb21073",
   "metadata": {},
   "source": [
    "Wordle 258 (solution `ahead`) is creating an issue with `adieu` - only 4 other words create the same pattern and they are riding the coattails of adieu and creeping into the top lists. The guess count is only 5, so this looks like a strong signal but it's driven by adieu, rather than the popularity of `advew`. I decided the simplest thing is to remove the 258 data point. Otherwise, all 4 of the other words show up in the top 15.\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "815da09b",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:31:52.962525Z",
     "start_time": "2022-03-22T00:31:52.907154Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>score_pattern</th>\n",
       "      <th>target</th>\n",
       "      <th>guess</th>\n",
       "      <th>score_frequency_rank</th>\n",
       "      <th>score_count_fraction</th>\n",
       "      <th>wordle_num</th>\n",
       "      <th>guess_count</th>\n",
       "      <th>commonality</th>\n",
       "      <th>weighted_rank</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>21010</td>\n",
       "      <td>ahead</td>\n",
       "      <td>adder</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.065498</td>\n",
       "      <td>258</td>\n",
       "      <td>5</td>\n",
       "      <td>888956</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>21010</td>\n",
       "      <td>ahead</td>\n",
       "      <td>adieu</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.065498</td>\n",
       "      <td>258</td>\n",
       "      <td>5</td>\n",
       "      <td>274747</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>112</th>\n",
       "      <td>21010</td>\n",
       "      <td>ahead</td>\n",
       "      <td>admen</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.065498</td>\n",
       "      <td>258</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>21010</td>\n",
       "      <td>ahead</td>\n",
       "      <td>advew</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.065498</td>\n",
       "      <td>258</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>133</th>\n",
       "      <td>21010</td>\n",
       "      <td>ahead</td>\n",
       "      <td>adzes</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.065498</td>\n",
       "      <td>258</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    score_pattern target  guess  score_frequency_rank  score_count_fraction  \\\n",
       "102         21010  ahead  adder                   3.0              0.065498   \n",
       "108         21010  ahead  adieu                   3.0              0.065498   \n",
       "112         21010  ahead  admen                   3.0              0.065498   \n",
       "130         21010  ahead  advew                   3.0              0.065498   \n",
       "133         21010  ahead  adzes                   3.0              0.065498   \n",
       "\n",
       "     wordle_num  guess_count  commonality  weighted_rank  \n",
       "102         258            5       888956            9.0  \n",
       "108         258            5       274747            9.0  \n",
       "112         258            5            0            9.0  \n",
       "130         258            5            0            9.0  \n",
       "133         258            5            0            9.0  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.query('wordle_num == 258 and score_pattern == \"21010\"')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "8330152c",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:31:52.986518Z",
     "start_time": "2022-03-22T00:31:52.963662Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1. adieu\n",
      "2. audio\n",
      "3. stare\n",
      "4. crane\n",
      "5. arise\n",
      "6. irate\n",
      "7. heart\n",
      "8. addio\n",
      "9. raise\n",
      "10. great\n",
      "11. train\n",
      "12. arose\n",
      "13. crate\n",
      "14. tears\n",
      "15. soare\n"
     ]
    }
   ],
   "source": [
    "for i, var in enumerate(top_openers.query('variable != \"guess_count\"')['variable'].head(15)):\n",
    "    print(f\"{i+1}. {var}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f177eb18",
   "metadata": {},
   "source": [
    "#### Just how popular is _adieu_?\n",
    "\n",
    "Ok, so _adieu_ is popular but exactly how many people are actually using it every day? To answer this, I looked at what fraction of all tweeted score patterns are consistent with `adieu` and plot that against how many other words could have also made that pattern."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "856d6399",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:31:53.073717Z",
     "start_time": "2022-03-22T00:31:52.987612Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "guess = 'adieu'\n",
    "df.query(f'guess == @guess and guess_count < 400').sort_values(\n",
    "    'wordle_num').plot.scatter(\n",
    "        x='guess_count',\n",
    "        y='score_count_fraction',\n",
    "    hover_data=['score_pattern','wordle_num'],\n",
    "       # color='guess_count',\n",
    "        title=f'{guess.upper()} popularity',backend='plotly',\n",
    "       # color_continuous_scale='bluered',\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7cc1bbc0",
   "metadata": {},
   "source": [
    "So in the real data, the minimum fraction that was `adieu` was about 4%.\n",
    "\n",
    "One could try and extrapolate the points and fit a line to this to see what it would like at a guess_count of 1 from the above data, but of course, we already fit a model to the data. The fitted Ridge model can predict what `adieu` (and all the other top choices) would do with a guess count of 1 and where `adieu` was the only word that could make it:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "d8b8fd0a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:34:17.784094Z",
     "start_time": "2022-03-22T00:34:17.598009Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>coef</th>\n",
       "      <th>variable</th>\n",
       "      <th>estimated_fraction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.017741</td>\n",
       "      <td>adieu</td>\n",
       "      <td>0.018581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.011133</td>\n",
       "      <td>audio</td>\n",
       "      <td>0.011972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.008327</td>\n",
       "      <td>stare</td>\n",
       "      <td>0.009166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.007077</td>\n",
       "      <td>crane</td>\n",
       "      <td>0.007917</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.006380</td>\n",
       "      <td>arise</td>\n",
       "      <td>0.007220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.006203</td>\n",
       "      <td>irate</td>\n",
       "      <td>0.007043</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.005777</td>\n",
       "      <td>heart</td>\n",
       "      <td>0.006617</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.005218</td>\n",
       "      <td>addio</td>\n",
       "      <td>0.006058</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.005135</td>\n",
       "      <td>raise</td>\n",
       "      <td>0.005975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.005063</td>\n",
       "      <td>great</td>\n",
       "      <td>0.005903</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.004999</td>\n",
       "      <td>train</td>\n",
       "      <td>0.005839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.004940</td>\n",
       "      <td>arose</td>\n",
       "      <td>0.005780</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.004466</td>\n",
       "      <td>crate</td>\n",
       "      <td>0.005306</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.004272</td>\n",
       "      <td>tears</td>\n",
       "      <td>0.005111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.004050</td>\n",
       "      <td>soare</td>\n",
       "      <td>0.004889</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        coef variable  estimated_fraction\n",
       "0   0.017741    adieu            0.018581\n",
       "1   0.011133    audio            0.011972\n",
       "2   0.008327    stare            0.009166\n",
       "3   0.007077    crane            0.007917\n",
       "4   0.006380    arise            0.007220\n",
       "5   0.006203    irate            0.007043\n",
       "6   0.005777    heart            0.006617\n",
       "7   0.005218    addio            0.006058\n",
       "8   0.005135    raise            0.005975\n",
       "9   0.005063    great            0.005903\n",
       "10  0.004999    train            0.005839\n",
       "11  0.004940    arose            0.005780\n",
       "12  0.004466    crate            0.005306\n",
       "13  0.004272    tears            0.005111\n",
       "14  0.004050    soare            0.004889"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "series_list = []\n",
    "for word in top_openers['variable'].head(15):\n",
    "    predict_dict = {word:1,'guess_count':(1 - guess_count_mean) / guess_count_std}\n",
    "    series_list.append(pd.Series({x:predict_dict.get(x,int(0)) for x in X.columns}))\n",
    "\n",
    "\n",
    "predict_on_this = pd.DataFrame(series_list)\n",
    "\n",
    "top_15 = top_openers.head(15).reset_index(drop=True)\n",
    "top_15.loc[:,'estimated_fraction'] = r.predict(predict_on_this)\n",
    "top_15"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "81947b5b",
   "metadata": {},
   "source": [
    "This indicates that `adieu` only represents 1.85% of all starters. The infamous wordle 258 implies it's more like 6%, but we've seen it less than 6% even with more possible words. It could be that wordle 258 was something of an outlier. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6ed77dc9",
   "metadata": {},
   "source": [
    "### CRANE\n",
    "\n",
    "On February 6, 3Blue1Brown released a video positing that [CRANE was the best opener](https://www.youtube.com/watch?v=v68zYyaEmEA). Though this was later [recanted](https://www.youtube.com/watch?v=fRed0Xmc2Wg), it generated plenty of [media](https://kotaku.com/wordle-starting-word-math-science-bot-algorithm-crane-p-1848496404) [coverage](https://www.forbes.com/sites/paultassi/2022/02/08/the-best-wordle-starter-word-a-first-guess-through-science/?sh=94c150a27909). CRANE was #3 on my list overall, how does it look before and after the video?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "826b6c74",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:24:17.910103Z",
     "start_time": "2022-03-22T00:24:13.088091Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ridge(alpha=10)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_recent = one_hot_encoded_data.query('wordle_num > 232')\n",
    "X= data_recent.drop(\n",
    "        columns=[ 'score_count_fraction', 'wordle_num']) # fit to the one hot encouded guesses and the total guess count\n",
    "y=data_recent['score_count_fraction'] # our dependent is the fraction of guesses that had the score pattern\n",
    "\n",
    "r = linear_model.Ridge(alpha=10)\n",
    "r.fit(X,y)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "82534db6",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:24:17.936339Z",
     "start_time": "2022-03-22T00:24:17.911031Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>coef</th>\n",
       "      <th>variable</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>0.016916</td>\n",
       "      <td>adieu</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>634</th>\n",
       "      <td>0.011274</td>\n",
       "      <td>audio</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10646</th>\n",
       "      <td>0.008455</td>\n",
       "      <td>stare</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2363</th>\n",
       "      <td>0.007766</td>\n",
       "      <td>crane</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5402</th>\n",
       "      <td>0.006561</td>\n",
       "      <td>irate</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           coef variable\n",
       "109    0.016916    adieu\n",
       "634    0.011274    audio\n",
       "10646  0.008455    stare\n",
       "2363   0.007766    crane\n",
       "5402   0.006561    irate"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(list(zip(r.coef_, r.feature_names_in_)),\n",
    "         columns=['coef', 'variable']).sort_values('coef',ascending=False).head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7156f510",
   "metadata": {},
   "source": [
    "CRANE is #3. What about before wordle 233?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "2943deea",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:24:19.903545Z",
     "start_time": "2022-03-22T00:24:17.937511Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ridge(alpha=10)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>coef</th>\n",
       "      <th>variable</th>\n",
       "      <th>guess_rank</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2363</th>\n",
       "      <td>0.000359</td>\n",
       "      <td>crane</td>\n",
       "      <td>1018.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          coef variable  guess_rank\n",
       "2363  0.000359    crane      1018.0"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_past = one_hot_encoded_data.query('wordle_num <= 232')\n",
    "X= data_past.drop(\n",
    "        columns=[ 'score_count_fraction', 'wordle_num']) # fit to the one hot encouded guesses and the total guess count\n",
    "y=data_past['score_count_fraction'] # our dependent is the fraction of guesses that had the score pattern\n",
    "\n",
    "r = linear_model.Ridge(alpha=10)\n",
    "r.fit(X,y)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "out = pd.DataFrame(list(zip(r.coef_, r.feature_names_in_)),\n",
    "         columns=['coef', 'variable']).sort_values('coef',ascending=False)\n",
    "out['guess_rank'] = out['coef'].rank(ascending=False)\n",
    "out.query('variable == \"crane\"')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "96e90e53",
   "metadata": {},
   "source": [
    "Prior to Wordle 233, CRANE ranked 1018! Quite the turnaround. (Alpha values may not be optimal for these smaller sample sizes) You can see this even in the cruder analysis, where the ranks for CRANE were much lower in the past."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "938148e0",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-03-22T00:24:21.075244Z",
     "start_time": "2022-03-22T00:24:19.904417Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "\n",
    "guess = 'crane'\n",
    "myplot = df.query(f'guess == @guess').sort_values(\n",
    "    'wordle_num').plot.scatter(\n",
    "        x='wordle_num',\n",
    "        y='score_frequency_rank',\n",
    "        color='guess_count',\n",
    "        title=f'{guess.upper()} popularity',backend='plotly',\n",
    "        color_continuous_scale='bluered',\n",
    "    )\n",
    "myplot.update_yaxes(autorange=\"reversed\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3ca27b2a",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.6"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "338.796875px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}