Rematch winners

[Fred]

In the Marietta thread I am just now having a chance to read, there was an a statement that in rematches the team that lost the first time almost always wins the second time.  At least based on GA high school stats this not true at all.  Loren Maxwell is a founding member of the Georgia High School Historians association and is a mathematician that produces the Maxwell ratings for GA.  I asked him this question in 2017 and here is his response.  You can argue perception all day long but here is hard statistical data that disproves the old perception.  I have the spreadsheet with his data if anyone wants me to post it.  Todd is Todd Holcomb who is/was in charge of high school sports for the Atlanta Journal Constitution and is one of the authors of the Georgia High School Football Daily newsletter that gets quoted here regularly.
**********************************************************************************************************

I've never done a list in my articles with that info, but I've generated that list a few times for Todd focusing mainly on rematches in the semifinals, finals, or something similar.

 

In total, there have been an unbelievable number of rematches -- nearly 4,000 since 1948 in the GHSA, the vast majority of which come from the old days when the first two rounds of the playoffs were among region opponents.

 

However, if we look only at quarterfinals and beyond, there have been a much more manageable 104 rematches, not counting this weekend's Brookwood-Colquitt County and Marist-St. Pius X rematches.

 

Of the 104, five were ties in the regular season, so we can toss those out, leaving 99 games to consider.
 

The original winning team in those cases is 61-38 (61.6%).  In the semifinals specifically, the original winning team is 29-15 (65.9%).

 

Note also that Aquinas and Stratford Academy played each other during the regular season twice in 2015 and then meet again in the semifinals for a total of three meetings in the same season.  Aquinas won all three games, so I only counted their last regular season game.

 

[dntn31]

7 minutes ago, Fred said:

In the Marietta thread I am just now having a chance to read, there was an a statement that in rematches the team that lost the first time almost always wins the second time.  At least based on GA high school stats this not true at all.  Loren Maxwell is a founding member of the Georgia High School Historians association and is a mathematician that produces the Maxwell ratings for GA.  I asked him this question in 2017 and here is his response.  You can argue perception all day long but here is hard statistical data that disproves the old perception.  I have the spreadsheet with his data if anyone wants me to post it.  Todd is Todd Holcomb who is/was in charge of high school sports for the Atlanta Journal Constitution and is one of the authors of the Georgia High School Football Daily newsletter that gets quoted here regularly.
**********************************************************************************************************

I've never done a list in my articles with that info, but I've generated that list a few times for Todd focusing mainly on rematches in the semifinals, finals, or something similar.

 

In total, there have been an unbelievable number of rematches -- nearly 4,000 since 1948 in the GHSA, the vast majority of which come from the old days when the first two rounds of the playoffs were among region opponents.

 

However, if we look only at quarterfinals and beyond, there have been a much more manageable 104 rematches, not counting this weekend's Brookwood-Colquitt County and Marist-St. Pius X rematches.

 

Of the 104, five were ties in the regular season, so we can toss those out, leaving 99 games to consider.
 

The original winning team in those cases is 61-38 (61.6%).  In the semifinals specifically, the original winning team is 29-15 (65.9%).

 

Note also that Aquinas and Stratford Academy played each other during the regular season twice in 2015 and then meet again in the semifinals for a total of three meetings in the same season.  Aquinas won all three games, so I only counted their last regular season game.

 

Interesting. I wonder what the data looks like if you only looked at rematches between teams where the original MOV was 14 points or less. I would think a rematch would only potentially favor the loser of the original game if the two teams are somewhat evenly matched. On the other hand, if one team is blowing out another team by 3+ TDs then I would expect a rematch to go the same way... 

[HawgGoneIt]

5 hours ago, dntn31 said:

Interesting. I wonder what the data looks like if you only looked at rematches between teams where the original MOV was 14 points or less. I would think a rematch would only potentially favor the loser of the original game if the two teams are somewhat evenly matched. On the other hand, if one team is blowing out another team by 3+ TDs then I would expect a rematch to go the same way... 

This makes a lot of sense. Adding in the "evenly matched" criteria certainly could have an effect on the numbers I'd imagine. 

[Wosinc]

10 hours ago, dntn31 said:

Interesting. I wonder what the data looks like if you only looked at rematches between teams where the original MOV was 14 points or less. I would think a rematch would only potentially favor the loser of the original game if the two teams are somewhat evenly matched. On the other hand, if one team is blowing out another team by 3+ TDs then I would expect a rematch to go the same way... 

Friggin’ common sense.  You don’t belong here.

[badrouter]

12 hours ago, Fred said:

In the Marietta thread I am just now having a chance to read, there was an a statement that in rematches the team that lost the first time almost always wins the second time. 

Since you responded to my post in the other thread, I'll assume you are referring to what I said there. What you've typed here is not what I claimed there. What I claimed was that the rematch almost always favors the team that lost the first time. This means the team that lost the first time figures to do noticeably better the second time around. It does not mean they will almost always win. One (of many, if you like)  example I gave in the other thread, for those who missed it, was Lakeland-Kathleen in 2018. Lakeland won the first match at Kathleen 50-0. The rematch at Lakeland was also won by Lakeland, comfortably and without drama. The score of that game was 29-7. Some would say there is no difference, as the same team won comfortably both times. I'm observing that the second game was a much better showing for Kathleen. The margin of defeat was reduced from 50 points to 22 points. Obviously, the difference in quality between the two teams was too significant for Kathleen to entirely reverse the outcome. Still, there was a clear effect seen. Almost entirely in the form of Lakeland scoring fewer points.

A rematch means increased familiarity of the teams. And this tends to favor the defenses. Maybe it's knowing the signals, the snap count, being more prepared and knowing tendencies, or some combination of those. You also see this a lot when teams scrimmage 1s vs. 1s, especially in spring and summer ball where the only opponent they face is their own team. 

Mater Dei-SJB the last two years also shows this. The 2018 pair of games is fairly obvious in this trend towards defense etc. But, even the 2019 game, at closer glance, shows this as well. The first game the teams combined for 321 rushing yards and only 1 INT. The second game, the teams combined for only 49 rushing yards and Bryce Young threw three picks. That there were still a lot of points scored and a lot of passing yards speaks to the rare quality of both QBs. They weren't typical QBs by any means.

[Fred]

8 hours ago, badrouter said:

Since you responded to my post in the other thread, I'll assume you are referring to what I said there. What you've typed here is not what I claimed there. What I claimed was that the rematch almost always favors the team that lost the first time. This means the team that lost the first time figures to do noticeably better the second time around. It does not mean they will almost always win. One (of many, if you like)  example I gave in the other thread, for those who missed it, was Lakeland-Kathleen in 2018. Lakeland won the first match at Kathleen 50-0. The rematch at Lakeland was also won by Lakeland, comfortably and without drama. The score of that game was 29-7. Some would say there is no difference, as the same team won comfortably both times. I'm observing that the second game was a much better showing for Kathleen. The margin of defeat was reduced from 50 points to 22 points. Obviously, the difference in quality between the two teams was too significant for Kathleen to entirely reverse the outcome. Still, there was a clear effect seen. Almost entirely in the form of Lakeland scoring fewer points.

A rematch means increased familiarity of the teams. And this tends to favor the defenses. Maybe it's knowing the signals, the snap count, being more prepared and knowing tendencies, or some combination of those. You also see this a lot when teams scrimmage 1s vs. 1s, especially in spring and summer ball where the only opponent they face is their own team. 

Mater Dei-SJB the last two years also shows this. The 2018 pair of games is fairly obvious in this trend towards defense etc. But, even the 2019 game, at closer glance, shows this as well. The first game the teams combined for 321 rushing yards and only 1 INT. The second game, the teams combined for only 49 rushing yards and Bryce Young threw three picks. That there were still a lot of points scored and a lot of passing yards speaks to the rare quality of both QBs. They weren't typical QBs by any means.

What you said was

On 12/17/2019 at 9:01 PM, badrouter said:

But the rematch almost always favors the team that lost the 1st time. And, regardless, that just means the teams split the season series. That they'd lose in any circumstances to that team by three touchdowns tells you EC isn't nationally elite.

Your second sentence implies in the second game the first games loser wins.  In the above response you proceed to cherry pick games that support your position.  Your sample size is too small.  I said that doesn't statistically hold true and provided an analysis from 99 games that says different.  Spin it how you want.

[badrouter]

1 hour ago, Fred said:

What you said was

Your second sentence implies in the second game the first games loser wins.  In the above response you proceed to cherry pick games that support your position.  Your sample size is too small.  I said that doesn't statistically hold true and provided an analysis from 99 games that says different.  Spin it how you want.

I can offer numerous more examples. I'll keep it short for now because it's Christmas. A few rematches in conference championship games also come to mind. This season, Ohio State beat Wisconsin by 31 in the first game, never trailing. In the rematch, Wisconsin led by two touchdowns at half before losing by 13. Much better showing for the 1st game loser. In 2017, Auburn hammered Georgia in game one, only to lose by 17 in the rematch, clearing the way to a Georgia playoff appearance. 

Below are many more rematches in college football. I can give more HS examples at a later time.

Year	First Game	Second Game	Result	Point Swing
2016	North Texas 35 — Army 18	North Texas 31 — Army 38	W	24
2012	Iowa State 38 — Tulsa 23	Iowa State 17 — Tulsa	W	29
2011	Alabama 6 – LSU 9	Alabama 21 – LSU 0	W	24
2010	Nebraska 56 – Washington 21	Nebraska 7 – Washington 19	W	47
2008	Navy 24 – Wake Forest 17	Navy 19 – Wake Forest 29	W	17
2008	Houston 28 – Air Force 31	Houston 34 – Air Force 28	W	9

• Average point swing: 25.0

 

REMAINING BOWL GAME REMATCHES

Year	First Game	Second Game	Result	Point Swing
2007	Purdue 45 – C. Michigan 22	Purdue 51 – C. Michigan 48	L	20
2007	UCLA 27 – BYU 17	UCLA 16 – BYU 17	W	11
2003	Miami 22 – Florida State 14	Miami 16 – Florida State 10	L	2
2003	West Virginia 7 – Maryland 31	West Virginia 7 – Maryland 41	L	-10
1997	Notre Dame 24 – LSU 6	Notre Dame 9 – LSU 27	W	36
1996	Florida 21 – florida state 24	Florida 52 – florida state 20	W	35
1995	Toledo 49 – Nevada 35	Toledo 40 – Nevada 37	L	11
1994	Florida 31 – florida state 31	Florida 17 – florida state 23	n/a	6
1994	UNLV 23 – C. Michigan 35	UNLV 52 – C. Michigan 23	W	41
1987	Michigan State 27 – USC 13	Michigan State 20 – USC 17	L	11
1982	Michigan 27 – UCLA 31	Michigan 14 – UCLA 24	L	-6
1978	Oklahoma 14 – Nebraska 17	Oklahoma 31 – Nebraska 24	W	10
1975	Ohio State 41 – UCLA 20	Ohio State 10 – UCLA 23	W	34
1965	Michigan State 13 – UCLA 3	Michigan State 12 – UCLA 14	W	12
1959	LSU 7 – Ole Miss 3	LSU 0 – Ole Miss 21	W	24
1956	Iowa 14 – Oregon State 13	Iowa 35 – Oregon State 19	L	-15
1945	Wake Forest 13 – South Carolina 13	Wake Forest 28 – South Carolina 14	n/a	14
1943	TAMU 28 – LSU 13	TAMU 14 – LSU 19	W	20

• Team that lost the first game: 15 - 7

• Average point swing: 17.5

[HooverOutlaw]

In 1999 Alabama beat Florida twice in Gainesville 40-39 and in the SEC title game 34-7.