Over in
@ELA 's simulation there is a plausible chance that Ohio State, Oregon, and Michigan could all end up tied at 8-1 and based on the league's tiebreakers that would be broken by SoS (cumulative conference record of conference opponents).
Within the scenario after the weekend of November 8 (so heading into the last three weeks), the cumulative records of those teams' opponents are:
Within this scenario the cumulative records of conference opponents for tOSU, Ore, and M are:- 26-30 Ohio State's opponents
- 25-31 Oregon's opponents
- 22-34 Michigan's opponents
Each team's nine B1G opponents will play nine league games so each team's opponents will end up with 81 total games (9*9) and these three each have 56 heading into the last three weeks so there are 25 remaining. Consequently, it seems like Ohio State's lead isn't all that big. Looking as compared to Michigan, the Buckeye's opponents are 4 games better. With 25 games remaining it seems like Michigan's opponents could make that up, they only have to be slightly better over those last 25 games, right?
As Lee Corso would say, not so fast my friend. That lead is a lot bigger than it seems for a number of reasons.
Michigan needs their opponents to make up 4 games in 25 but in order for this tie to happen Michigan has to beat Ohio State. That obviously helps Michigan as it gets them into the tie but counter-intuitively, it actually hurts Michigan in the tiebreakers because Michigan is an Ohio State opponent and Ohio State is a Michigan opponent. Thus, Michigan beating Ohio State makes it now 5 games in 24 that Michigan needs their opponents to make up.
Next, in order for this tie to happen both Ohio State and Michigan need to win their other two games so that makes it five games in 20 that Michigan needs their opponents to make up.
Next we have common opponents. UW, UDub, and PU are opponents of both tOSU and M so their 3, 3, and 2 remaining games respectively will provide no advantage nor disadvantage vis-a-vis tiebreakers. Thus Michigan now needs their opponents to make up 5 games in 12.
Those remaining games are:
For Michigan's opponents:
- UNL at PSU
- UNL vs IA
- USC vs IA
- USC at Ore
- USC vs UCLA
- UMD at IL
- UMD vs MSU
- MSU vs PSU
- MSU at IA
- MSU vs UMD
- NU vs MN
- NU at IL
The games in
bold are two sides of the same game. Collectively, Michigan's opponents are guaranteed to go 1-1 in this game since they are on both sides.
For Ohio State's opponents:
- MN at Ore
- MN at NU
- MN vs UW
- IL vs UMD
- IL at UW
- IL vs NU
- PSU at MSU
- PSU vs UNL
- PSU at RU
- UCLA vs UW
- UCLA at USC
- RU vs PSU
The bottom line is that Michigan's opponents have 10 and Ohio State's opponents have 12 remaining games that make a difference for this tiebreaker and Michigan needs their opponents to gain 5 in those 10/12 games. It isn't mathematically impossible but it is highly unlikely.
NOTE:
I really don't mean this post to be about tOSU/M. That isn't the point, they are just the example that came up due to the situation in ELA's simulation. The point is a larger one: If any given team's opponents are down by 4 or 5 games late in the season it might seem like their opponents have a pretty good chance of catch up but realistically, they probably don't. Four or 5 games doesn't seem like a lot out of 25 but when you start working through the possibilities you end up needing more like 5 in 10 and that is extraordinarily unlikely.
