suggestion: best games list « Forums

Forums | FanGraphs Baseball » Site Suggestions

suggestion: best games list

(13 posts)

jkent
Member

After looking at A-rod's stellar performance last weekend, I wondered who else had such an impact on one game.
I think a list of the best games played each year (in terms of WPA) would be a good addition. It could be a rank of the highest WPA for a given game. One for the worst games each year is a possibility as well.
Anway, this site is great! Thanks!

Posted 2 years ago #
David Appelman
Founder

A best games and best plays leaderboard would be nice to have. I'll see what I can do about it. Maybe by the end of the month.

Posted 2 years ago #
Dumonk
Member

In regards to this, could this also be done on a team level as well?

Posted 2 years ago #
tangotiger
Member

At the team level, every team is +.500 or -.500 each game. Are you asking for something else, like team's batters, team's bullpen, etc?

Posted 2 years ago #
dackle
Member

There was talk last year (http://www.fangraphs.com/forums/showthread.php?t=85) after the Dodgers' 4-HR comeback of ranking games according to various metrics -- highest cumulative WPA, lowest standard deviation throughout the game etc. Any chance of incorporating these into a leaderboard?
Also, the ability to sort all of an individual player's plate appearances by WPA is pretty nifty. Are there any plans of sorting by team or even league (ie what were the 10 biggest plate appearances by WPA in the majors this season?)

Posted 2 years ago #
Dumonk
Member

tangotiger;770 At the team level, every team is +.500 or -.500 each game. Are you asking for something else, like team's batters, team's bullpen, etc?
I was mainly asking, it be nice to go to the TEAM page and see the top 10 (could be any number really) performances of batters per game. Same for starting pitchers or relief pitchers for that matter.

Posted 2 years ago #
dackle
Member

Hi David, are there any plans to add a "best games" list to the site? Just a page with:
- visitor - home - score - innings - average leverage index - total win advancements - lowest WPA at any point in the game for the winning team (ie to highlight the best comebacks)
Regarding what makes a good game ... I have been tracking average LI and it's an excellent way of picking archived games to watch on MLB-TV. In the discussion last year re: the Dodgers' four home-run game you listed some of the highest-ranked games by various metrics. Average LI came up with a lot of extra-inning games, but one way around this is to "detrend" the average LI with a linear equation -- eg (innings played * .222) -1.04. A nine-inning game would have an expected average LI of 0.96, a 12-inning game 1.62. A list of games ranked by actual average LI minus expected average LI generates a nice list of 9- and 10-inning nailbiters. I think Studes is of the view that a great game is a "wild" game, with a high total of win advancements. I agree to a point, but having watched a few of those 15-14 games that last four hours, I'd have to say that a shorter affair with lots of high-leverage at-bats is more entertaining (eg http://www.fangraphs.com/wins.aspx?date=2007-05-07&team=Cardinals&dh=0&season=2007) or (http://www.fangraphs.com/wins.aspx?date=2007-04-05&team=Yankees&dh=0&season=2007 with the snow falling in the eighth and ninth innings), although, you can use average LI of each half inning to edit a four-hour game down to an hour and a half of its most important half innings. But, I think there's almost a 1:1 relationship between leverage index and tension, and when there is tension we sit forward in our easy chair and pay closer attention to the game. Whereas I associate total win advancements with a feeling of bored admiration (ie somewhat sarcastically asking your neighbour: "Wow, can you believe this game, they were down 11-6 and now it's 13-12"). Fun, but the entertainment level isn't quite so high (when you realize the game started four and a half hours ago).
I noticed also that most of the games with the highest average LI are played in the National League ...

Posted 2 years ago #
tangotiger
Member

Dackle,
Interesting idea to compare the LI of the game based on the number of innings played.
I don't really see it as a problem that you'll have a 14-inning game "winning" the average LI contest. Extra inning games are great! But, if you want to compare a game, over and above its "extra inning drama", that's ok.
I think that cumulative absolute WPA minus .0346*PA probably would work well. The ".0346" refers to the average win probabilty change per PA.

Posted 2 years ago #
dackle
Member

Yeah, they are great, I agree. I suppose I'd rather watch a nine-inning game with an avg LI of 2.00 than a game with the same avg LI but spread out over 14 innings, I guess because the tension is compressed within a shorter timeframe.
But in terms of what makes a good game, average LI is probably closer to the mark than total win advancements. I've tried watching the Cubs/Phillies 23-22 game ... I get bored. The White Sox/Brewers 25-inning game doesn't interest me (except maybe the last half inning). The Yankees/Red Sox 4h 45m 14-11 nine-inning game doesn't do it for me either. Seems like all the best games have a high average LI, are played in under 2h 30m, and preferably in front of a large crowd in an atmospheric stadium with threatening weather between contenders with marquee players. I realize that not all of that is programmable, but even just a list of highest average LI would be a great starting point!

Posted 2 years ago #
tangotiger
Member

I'm certainly comig around to you.
For example, the best overtime game I saw was the NYI/WAS game back in 1987 or so. Triple-OT, or was it quadruple? All I know is that I went to bed at 2 AM.
I suppose that the longer the game, the more drama there better be. Your idea makes alot of sense.
Another great one in the world of tennis was John McEnroe in ... I don't remember... the one where there was back-and-forth with Aaron Krickstein (or however you spell his name). If you are going to get alot of volley back and forth, each one better be more tense than the next, and they definitely delivered.
So, I think the length-component makes alot of sense.

Posted 2 years ago #
dackle
Member

I hate to keep bugging David about a best games list, but ... would be cool to see where last night's Blue Jays comeback stacks up against other comebacks this year -- Devil Rays were up 11-6 with a runner on first and none out in the top of the ninth -- WE of 0.6%.

Posted 2 years ago #
tangotiger
Member

Rereading my post. I can't believe I said McEnroe and not Jimmy Connors.

Posted 1 month ago #
JoeR43
Member

Think aLI is a good barometer to how good a game actually was?

An examples of a huge aLI games:
http://www.fangraphs.com/wins.aspx?date=2004-10-17&team=Red%20Sox&dh=0&season=2004

of course the problem with this is that it doesn't factor in miraculous comebacks.

Posted 1 month ago #