Those Four Charts
I've finally gotten around to writing them up. And added two more.
At the top, let’s take a moment to recall and celebrate the great Alex Trebek, on the fifth anniversary of his passing. An incalculable loss to Jeopardy!, to be sure. From wherever he is now, he surely continues to watch over and celebrate our beloved game, as it carries on and moves forward — just as he hoped for and intended.
Several weeks from the end of Season 41, there was some chatter about how the percentage of correct responses in Final Jeopardy! was lower than in the few preceding seasons. J! Archive has stats pages on this for each season. Since the most recent game not in the Archive is from January 1997 (Season 13), I charted out the percentages of Final Jeopardy! gets in the previous twenty-eight seasons, for regular play only.
The percentage came up a couple of points in the closing stages of the season. But at 42.26%, it was still the second lowest in the time span; only 1998-99 saw a lower number. I’d need to see at least one more season of Final running at least that difficult before endorsing any conscious shift in intent by the writers.
Season 42 through 45 games: 40.91% (54/132)
The Archive also has the same data for Daily Doubles:
Here, the figure bounced back up to where it usually is, after dipping in Season 40. There are more significant outliers on both the high and low sides here — in the former case, Season 35.1
Season 42 through 45 games: 71.11% (96/135)
Looking at those pages, I also noticed they show the mean clue at which each of the three Daily Doubles are found. And this is where you really see an effect, and its likely cause is pretty obvious.
Season 42 through 45 games:
Jeopardy! Round Daily Double — 10.24
First Double Jeopardy! Round Daily Double — 7.42
Second Double Jeopardy! Round Daily Double — 15.62
Looking at the first round, the time it took to find the Daily Double was pretty constant throughout the 2000s and 2010s, but it has fallen to new “lows” (that is, earlier) in each of the last four seasons. In Double Jeopardy, the story is slightly different. There was a slight downward trend throughout the 2010s — but again, these all-important clues are being found earlier season after season in the 2020s.
This seems a pretty direct effect of James Holzhauer’s play. Playing Jeopardy! categories top-down, as had been done for so long before 2019, is absolutely passé in the mid-2020s. I’m reminded of images by Kirk Goldsberry2 of the most preferred shot locations in the NBA two decades ago and now; they show that the mid-range jumper has all but disappeared, in favor of three-pointers and shots in the paint.
As with the Final Jeopardy! correct response rate, I’m not convinced we’ve reached the point where compensatory action is necessary. How the show might deal with this if it wanted to? Two possibilities have been floated out:
Allow players to select category only, giving them the lowest value clue remaining (effectively “forcing” top-down play)
Equalize the likelihood of finding a Daily Double among the five rows of the gameboard
I’m much more partial to the second option. If players are going to play the top rows last, then making it more likely that Daily Doubles appear there will result in their being found later, reversing the trend over the last half-decade. And if each row is equally likely to contain a Daily Double, a deliberate choice to start on one level of the board won’t matter. Taking away contestants’ choice of row (and thus, clue difficulty) is too drastic an alteration; it has a feeling of trying to change the game too fundamentally. Plus, players may want to attack the board in different ways for reasons other than Daily Double hunting. Their ability to do so would be severely restricted by forcing top-down play, and those variations in strategy shouldn’t be denied to them.
But it seems the show doesn’t see this as problematic. I’ve drawn up two more charts to demonstrate this. Here are the Daily Double distributions for the Jeopardy and Double Jeopardy rounds from 1997-98 to 2024-25:
Not much change over the long time horizon, indeed. The only shift of any real note on these two images is the disappearance of the small gray area at the top, which represents unrevealed Daily Doubles. The last time a round ended with a Daily Double still out there was October 28, 2020 — more than five years ago. I don’t recall the show ever mentioning there now being a requirement that every Daily Double be played. But given the changes implemented over the last four years, it would make sense. Eliminating unplayed Daily Doubles ensures that those key game aspects are equalized for all players — something critically important to comparisons between them in consideration of invitations back for Second Chance, and now, Champions Wildcard. (It’s analogous to the requirement to play all clues in the quarterfinals of fifteen-player, ten-game “traditional format” tournaments.)
Excluding James Holzhauer from the 2018-19 numbers yields a Daily Double get rate of 69.84%. The equivalent number for Final is 51.63%.
Specifically in that article from The Ringer, image 2. Image 3, showing points per shot over the last twenty-five seasons, takes things further.