The Phillies lost to the Dodgers on Tuesday with the final play having the two base runners on 2nd and 3rd getting tagged out resulting in a 0.558 Win Probability Added for the Dodgers. Where does this defensive WPA rank all-time?

With 1 out in the bottom of the 9th the Dodgers were ahead 2-1. Phillies had runners on 2nd & 3rd. The batter hit a ground ball to the 3Bman. He throws home as the runner on 3rd went home.

But the runner ran back to 3rd seeing that the catcher had the ball. The catcher runs him back to 3rd. But Betts ran in from SS. He took a throw from the catcher and tagged out the runner.

By this time the runner on 2nd had run to third but then ran back to 2nd base when he saw the original runner retreat to 3rd base. 

As soon as Betts tagged the original man on 3rd he turned and threw the ball to the 2Bman who tagged the runner trying to get back. Game over.

 

Here is a video 

 

Here is a video posted by Todd Zolecki 

 

This link from Stathead has the results from the search For year 2026, Involving Philadelphia Phillies, Game-Ending, Sorted by Championship Win Probability Added.

 

If you sort by WPA, this play is the highest listed.

 

This link from Stathead has the results from the search In years 1903 to 2026, requiring WPA >= .5, Sorted by Win Probability Added. 

 

I had selected "non-SO outs->(to Pos) as the Event Type. I also used the "All Outs" option under "Double/Triple Plays." Here are the top plays:

 

Rk	Date	WPA	Bat Tm	Def Tm
1	7/15/1930	0.572	PHI	STL
2	9/25/1948	0.571	BSN	NYG
3	9/10/1973	0.571	BAL	BOS
4	4/8/2018	0.57	NYY	BAL
5	5/1/2000	0.57	CLE	NYY
6	8/4/1921	0.568	BOS	CHW
7	5/11/1997	0.566	MIN	TOR
8	9/24/2000	0.566	BOS	BAL
9	7/1/1939	0.564	STL	PIT
10	8/27/1996	0.564	COL	CIN
11	9/13/1982	0.564	BOS	CLE
12	9/16/1930	0.564	BRO	STL
13	7/2/2021	0.563	OAK	BOS
14	7/21/1999	0.558	CLE	TOR

 

So it looks like this play by the Dodgers ties for the 14th highest WPA ever in the Stathead databse by a defensive team. Also, the search results will allow you to go to to the box scores of the games and it also tells you what the play was and all the relevant situational details. 

 

This most recent play from the Dodgers-Phillies game did not show up on these search results. I asked BB Ref about this and they told me 

"It looks like the event is coded as a "Fielder's Choice", and for some reason, that doesn't show up when filtering for "Non-SO outs". That seems like an error in the codebase, so I'll have our engineering team look into it." 

Also, the actual top 4 results you will see if you click on the link are WPAs greater than one (which makes no sense). I omitted those in the table above. BB Ref told me that "I'll have our engineering team look into the problem" about this issue as well. The 5th one is actually a play in which a run scored on an error on the final play of the game and the batting team won.

Interesting Game Score trends

There are 52 pitcher starts with an 85+ Game Score in 7 IP or less in the Stathead database (earliest is 1993). 51 of them are since 2000, 40 are since 2016. 4 this year, all since 6-10. 7 had 0 hits & all of those are since 2016.

https://www.sports-reference.com/stathead/tiny/PBBSM

 

There are 118 games when the starter pitched 6 or 7 innings with no hits allowed in the Stathead database. 80 since 2010. 69 are from 2017-2026

https://www.sports-reference.com/stathead/tiny/sGV5l

 

Stathead has 170 games when a starter had a Game Score of 62+ in 4 IP or less. 142 of them since the year 2000. 14 this year. Only 14 before 1980

https://www.sports-reference.com/stathead/tiny/baBIe

What happens in games when one team has a higher OPS and the other has a higher AVG?

I don't recall seeing this specific question raised before. So I asked bases.chat, the AI program created by fellow SABR Ted Ingram. If there is a way to search this easily on Stathead or somewhere else I don't know what that is.

In almost every year I looked at, the teams that have a higher OPS than their opponent in games when the other team had a higher AVG win at least 66% of the time and sometimes over 70%. The lowest year was just 64%. 

I had it check several different years (At first I just did 2021-25 but Jessica Brand suggested I look at earlier periods to see if there was a trend). The table below has the results. After that will be a table summarizing the three periods.

Season	Divergent Games	Wins	Win %
1982	298	218	73.20%
1983	270	185	68.50%
1984	270	183	67.80%
1985	279	197	70.60%
1986	304	203	66.80%
2001	321	206	64.20%
2002	313	212	67.70%
2003	335	229	68.40%
2004	328	233	71.00%
2005	344	227	66.00%
2021	311	218	70.10%
2022	313	225	71.90%
2023	347	239	68.90%
2024	289	205	70.90%
2025	301	203	67.40%

Divergent games just means how many such games happened in each year. 

Now the summary table.

Season	Divergent Games	Wins	Win %
1982-86	1421	986	69.39%
2001-05	1641	1107	67.46%
2021-25	1561	1090	69.83%
Total	4623	3183	68.85%

Earlier I said I did not know how to search this without bases.chat. But I did ask it which team in 2025 played in the most games like this and it said it was the Orioles and that they were 14-3 in those games. 

I called up the batting game logs for the Orioles at Baseball Reference and then used Stathead to find the same logs for their opponents. By sorting, matching games and finding differentials I could see that the Orioles were 14-3 in such games. So that helps show that bases.chat was doing this right.