Is overtime really that random?

Submitted by snarling wolverine on

Michigan is now a very impressive 10-2 in overtime games.  The games:

2000 Orange Bowl vs. Alabama - win

2002 PSU - win

2004 MSU - win (3 OT)

2005 at MSU - win

2005 at Iowa - win

2009 at MSU - loss

2010 Illinois - win (3 OT)

2012 Sugar Bowl vs Virginia Tech - win 

2012 Northwestern - win

2013 at PSU - loss (4 OT)

2013 at Northwestern - win (3 OT)

2015 at Indiana - win (2 OT)

 Lloyd Carr went 5-0 in overtime; RichRod 1-1; Hoke 3-1; and Harbaugh is now 1-0.  By every measure we've done well in overtime.  We've never lost in it at home (4-0) or at a neutral site (2-0), and even on the road we're still 4-2.  When it goes beyond the first session, we're 4-1.  

Should we assume OT is actually a 50-50 proposition and that Michigan has been extremely fortunate?  I'm not sure.  I don't have the numbers for other teams, but just in general when I'm watching, it seems like the favored team usually ends up winning.  I think OSU has an overtime record similar to ours, for instance.  I would be curious to see what the records are for more teams.

MGoBoard
Football

FauxMo

November 15th, 2015 at 3:30 PM ^

OT, it seems to me, favors the team with the better defense. I'd be curious to see what our defensive rankings were compared to the opponent in each of those games. I'll bet we had the better defense in most of the OT games we won.

 

FYI, this is just a hunch, not based on research or anything...

Joined: 06/17/2010

MGoPoints: 137521

wahooverine

November 15th, 2015 at 5:34 PM ^

Agree with this but alternative view is that you can think of OT as a mini game wherein the team favored before the game ought to be favored as well for the same reasons.  Because OT is less time and more structured there is less potential for chaos to play a role, the same chaos that over the longer timeline of regulation, presumably played a role in the underdogs playing to an even score.

Joined: 11/14/2011

MGoPoints: 6799

Red is Blue

November 16th, 2015 at 2:56 PM ^

Sure, they're crucial, but does one team really have an advatage over the other?  Given you start at the 25, if you don't lose yards it is a 42 yarder.  Certainly, some FG kickers are more likely to make from that distance, but I would think the make percentage average for a P5 fg kicker from 42 and in probably isn't a great spread.  Even moreso for extra points.  The advantage you have in FG kicking, would be much more pronounced a littler further out.  If you're FG kicker can make 55 yarders, you've probably got an advantage over the other team.  But in OT, a longer kicking FG advantage rarely manifests itself. 

Joined: 05/13/2010

MGoPoints: 23889

Heteroskedastic

November 15th, 2015 at 4:48 PM ^

The only time I have ever heard of overtime being a 50/50 proposition was when discussing NFL games. But that assumes greater parity among a smaller pool of teams and the OT format itself makes for a more equal proposition, especially before the recent modifications. . I have never heard NCAAF OT referred to as a even proposition. I would venture to say a teams OT winning percentage is highly correlated to it overall winning percentage during that same period.

Joined: 08/23/2014

MGoPoints: 495

Heteroskedastic

November 15th, 2015 at 6:35 PM ^

I am having issues posting the scatter, but the correlation between OT Win % a Overall Win % between 1996 and 2014 is about .32 (R-square~=.105). Not huge, but definately a relationship. I ran a simple regression of OT Win % on OT Opponent Win % and Own Overall Win %. The p-value of Own Overall Win % was .000006 which implies that overall win % almost certainly affects OT Win %. Using just OLS with the two independent variables to estimate the effect of Own Overall Win % on OT Win %, on average, a 10 ppt increase in overall win % increases OT Win % by 5.5 ppts holding all else constant.

Joined: 08/23/2014

MGoPoints: 495

Muttley

November 15th, 2015 at 11:01 PM ^

to a 50/50 proposition than a full game.

In college football OT, there are far fewer number of series than in a full game.

Consider the chances of a typical small-market MLB team against a$200 million salaried Yankees team in a single game versus over a 100 game season (to make the numbers easier.)  Let's assume that the Yankee's win 55% of their games and the small market team wins 45% of their games, and there is nothing unusual about the Yankees-small mkt team matchup.

After one game against each other, the Yankees are 55% likely to win while the small market team has a 45% chance of winning.  But in the 100 game season, the Yankees are likely to have won areound 55 games while the small market team is likely to have won around 45 games.  The chances that the small market teams win total will be above the Yankees are no where near 45%.

This is why the Yankees can usually buy their way into the playoffs with a talent-loaded team, but once they get to the playoffs, winning the seven (or five) game series becomes much less of a very likely thing.

Joined: 07/07/2009

MGoPoints: -73245

America

November 15th, 2015 at 4:35 PM ^

In a poker context, things similar to "OT is 50-5" are used as estimates to make calculations easier. It make sense to me that one team will have some small edge over the other, and literally almost nobody is going to be >60% to win before first snap, with the vast majority of teams being within a percentage or 2. Thus, in the interest of simplification, you could just call OT a 50-50 proposition when making strategic decisions in regulation as it will be close enough for that purpose. Just a thought.

Joined: 09/19/2009

MGoPoints: 812

kb

November 15th, 2015 at 3:33 PM ^

the games go on forever and the scores end up being 84-81 after six overtimes. They should have them start at the 50 yards line or something, not the 25.

Joined: 10/31/2009

MGoPoints: 7151

snarling wolverine

November 15th, 2015 at 3:43 PM ^

The really marathon OT games mostly stopped after they put in the requirement that teams have to go for two starting in the third session.  Of Michigan's 15 overtime games, 10 were resolved in the first session and 14 by the third.

If the NCAA wants to end games sooner than they are, they could require teams to go for two in the first or second OT.  To start each possession further back than the 25 might be counterproductive - you'd probably have more empty possessions, and thus more OT sessions.

 

 

 

Joined: 12/14/2011

MGoPoints: 97906

LSAClassOf2000

November 15th, 2015 at 4:58 PM ^

As I recall, a great majority of the really long OT affairs happened back about 15 years ago, with Arkansas being involved in two games that went to a 7th overtime. Tennessee was involved in one or two of those as well, but I cannot recall anything like that lately. The longest in recent years  that I can remember - and I could be missing something - is 4OT (didn't Buffalo have one go to 5OT in 2013 though?) and of course we were involved in one of those. 

Joined: 01/07/2011

MGoPoints: 130505

McSomething

November 16th, 2015 at 3:17 PM ^

Not true, the 2003 Arkansas/Kentucky game that went to 7 OTs operated under the "must go for 2" rule starting in the 3rd OT. As did the Arkansas/Ole Miss game in 2001. If you look at the box score, both teams missed a 2 point try in the 3rd OT. If PATs were still allowed, Ole Miss simply could've kicked it to win instead of attempting a 2 point conversion of their own. https://en.wikipedia.org/wiki/2003_Arkansas_vs._Kentucky_football_game#… https://en.wikipedia.org/wiki/Arkansas%E2%80%93Ole_Miss_football_rivalr…

Joined: 09/29/2010

MGoPoints: 36817

Carcajou

November 16th, 2015 at 2:37 AM ^

Start them farther back and it will be more difficult to score, meaning more periods. If anything, start them at the 20. That means (barring defensive penalites), the offense cannot gain more than one first down.

I've thought that the score in overtime should be listed after a decimal 28.14 to 28.07, or something to put things in better perspective. 

[For that matter, it seems to me OT wins and losses should be treated differently when speaking about conference tie-breakers, playoffs, and such]

Joined: 02/26/2009

MGoPoints: 5749

Carcajou

November 16th, 2015 at 2:46 AM ^

Another way to cut back on overtimes is to give a nudge or an incentive to resolve the game in regulation.

e.g. No coin flip to begin OT (unless the game ended 0-0).  The team that scored LAST, gets 2nd choice; the other team (that had the lead before they were tied), would get 1st choice. 

A team rallying late, under current rules has a 50% chance of winning the coin toss.  If they knew that merely tying to score would mean the other team would effectively win the toss, then going for 6 versus 3; or 2 verus 1- would become slightly more desirable.

Joined: 02/26/2009

MGoPoints: 5749

snarling wolverine

November 16th, 2015 at 11:13 AM ^

Interesting idea, but I don't know if it would make that big of a difference. If you're down 3, I don't think that would be enough to cause a coach to forgo a field goal and go for a long shot TD. I think they should require teams to go for two in the second OT session. Keep the PAT part of it for one session, but then bring in the conversion to wrap things up. Few games make it past 3 OTs now, and few would make it past two in this case.

Joined: 12/14/2011

MGoPoints: 97906

somewittyname

November 15th, 2015 at 3:41 PM ^

I have a theory, although I think it's more applicable to college hoops, that the better team usually wins in overtime. It often takes the underdog so much energy to just get to overtime that when it actually starts and momentum has a chance to refresh that the better team can then pull away. I think this holds reasonably true for college football.

Joined: 05/15/2009

MGoPoints: -2510

MGoDillon

November 15th, 2015 at 3:45 PM ^

Definitely agree. Seems pretty accurate, with a few exceptions here and there. That's why you see under dog teams, so often, go for the 2 - point for the win if it's applicable rather than take their chances by prolonging the game.



Sent from MGoBlog HD for iPhone & iPad

Joined: 08/17/2015

MGoPoints: 1454

Mr. Yost

November 15th, 2015 at 4:02 PM ^

...but in the end, I thought Wilson would kick the XP and he confirmed this after the game.

IU was running all over us. It's not like they were hanging on and they needed to get 2 and end it.

For example, when we played OSU in 2014. We were barely hanging on in that game. We had our offense on the field and they were the only thing working. We were gassed and it felt like we were losing grip of the game. The better team was asserting itself.

It was a shit playcall (should've moved the pocket)...but it was the right decision.

-------------------------------------------------

Sidenote...I've always had this rant. If you're EVER in the position on whether or not to go for it. Go for it (assuming you have a decent kicker) and run an illegal pick play.

If the refs call it...fine, you kick and play on. If they don't, you get a cheap 2-points for the win.

If you NEED the 2 points to tie...never run a pick play. If they call it, you're screwed.

Joined: 07/25/2011

MGoPoints: -9898220

MotownGoBlue

November 15th, 2015 at 4:32 PM ^

Our basketball team needs to take a page from this book. 2014-15 was an OT nightmare. Edit: With the exception of 2014-15, the Hoops team appears to be a good bit over .500 under Beilein.

Joined: 08/09/2014

MGoPoints: 25319

Mr. Yost

November 15th, 2015 at 3:57 PM ^

We were the better team or at least playing better in the majority of the games we went to OT. I'd be interested to see if we were favored in the majority of the games, which I assume we were.

We should've won the PSU game as well...but Borges single-handedly lost that game for us.

I don't recall any where were flat out lucky to win or we were the underdog, got to OT and pulled out the W.

Joined: 07/25/2011

MGoPoints: -9898220

snarling wolverine

November 15th, 2015 at 4:08 PM ^

We were lucky against Va Tech - we pretty much stole that one.

Bama ended on a fluke (missed PAT by their kicker) although they couldn't stop Brady at all, and once we unleashed him we were on fire.  

The two Northwestern OT games we were pretty fortunate - getting the Hail Mary in '12 and then the last-second kick.  I'm not sure we qualified as the underdog either time though.

In Braylonfest, we did get that onside kick which was pretty critical.  But we really should have never fallen behind so far in the first place.  Like the Bama game, once we started really attacking their secondary with our stud WRs, we couldn't be stopped.

 

 

Joined: 12/14/2011

MGoPoints: 97906

Mr. Yost

November 15th, 2015 at 4:21 PM ^

Not necessary how we got there...

I'll give you Bama, but I think Michigan still wins that game.

The VT game was the game I thought of as a rebuttal, but was too lazy to bring it up. It was NOT a catch...but we still should've lost that game if you're talking about which team played better.

Joined: 07/25/2011

MGoPoints: -9898220

NittanyFan

November 15th, 2015 at 4:02 PM ^

the data is available on Wikipedia (thru the end of the 2014 season):

https://en.wikipedia.org/wiki/NCAA_Division_I_FBS_overtime_records

Ohio State is at 8-2, so they're well above .500 also.  But on the significantly under-.500 side, you have teams like Alabama & USC.  At a glance, I don't see any obvious trend for "good" teams to tend to do better than .500 when they get to OT.

FWIW, Michigan at 10-2: if you flip a coin 12 times you'd get heads 10+ times only ~2% of the time.

Joined: 12/04/2012

MGoPoints: 42810