A Quick Look at Brady Hoke After Bye Weeks
I haven't had the time to write a diary this time around due to classes but I still thought it'd be interesting to see how Brady Hoke's teams have done coming off of a bye week.
Bottom line: Brady Hoke is 6-4 coming off of bye weeks. That's a .600 winning percentage. The national average is .480 and the Big Ten average is .350. As can be seen, Brady Hoke is outperforming both the national and Big Ten averages.
He was 6-2 coming off of bye weeks at Ball St. and he was 0-2 coming off of bye weeks at San Diego St.. Interestingly, in his first year at both SDSU and at Ball St. he lost the games coming off of a bye week.
Caveats abound regarding a small sample size, different conferences, quality of athletes, quality of opponents, etc.
P.S. If I have time I may write all this into a Diary later today, but I don't know how likely that will be.
October 28th, 2011 at 5:15 AM ^
Another caveat, both of those losses at SDSU came at the hands of BYU. 2009 was home against BYU and 2010 was @ BYU.
So, make of this what you will, but I'm still pretty confident about tomorrow.
October 28th, 2011 at 11:15 AM ^
That 2010 loss to BYU was the one where SDSU got jobbed on the replay challenge. Replay clearly shows fumble before down-by-contact but the replay booth claimed not to see that particular camera angle. Of the guys in the replay booth were BYU alumni and employees. Go figure.
1:45 mark:
October 28th, 2011 at 1:29 PM ^
so we've got that going for us.
October 28th, 2011 at 7:05 AM ^
When I saw this topic on the board for some reason I immediately read it "A quick look at lloyd brady after bye weeks". I'm also perplexed by the fact that I was disappointed when I oppened the thread and found that there was actual meaningful content.
October 28th, 2011 at 8:55 AM ^
I too was disappointed as I was expecting to see a picture of Coach Hoke pointing at something.
October 28th, 2011 at 7:09 AM ^
I love how week talk up the positives of coming off a bye week when we're the team who just had one, but dismiss the bye week effect when our opponent is the team coming off the bye
October 28th, 2011 at 8:06 AM ^
Its Michigan , we do that with everything. For gods sake.
October 28th, 2011 at 9:47 AM ^
For God's sake. fergodsakes.
Speak Michigan.
October 28th, 2011 at 9:48 AM ^
(or really anyone around here). The OP simply stated Hoke's record against the averages. I, myself, think the benefits and disincentives basically cancel one another out. That's why the national average is around .500. I also think most teams/conferences set up bye weeks before more difficult teams or rivals. The B1G obvious set up MSU's for before Michigan and Michigan's after MSU because of that game. LSU and 'Bama obviously have a bye week this week because they face each other next week. This leads me to believe that all else being even it is really a wash, there are positives and negatives.
October 28th, 2011 at 11:11 AM ^
If you're saying that the B1G tries to set up bye weeks before big games, why is it obvious that it relates to MSU the week before UM and UM after MSU? Are you suggesting that the B1G was trying to make it more difficult for us? That Purdue is a big game for UM? The B1G hates UM and Purdue? Please e'splain.
October 28th, 2011 at 11:53 AM ^
I was trying to say that the Big Ten usually sets up the bye weeks either directly before or directly after big games because they understand that those games are more taxing on players. I think its pretty obvious that the Big Ten isn't just picking them randomly.
Bye week for each team:
- Illinois: After PSU, before Michigan
- Indiana: After OSU, before MSU
- Iowa: Before PSU
- Michigan: After MSU
- Michigan State: After OSU, before Michigan
- Minnesota: Before Neb.
- Nebraska: After OSU
- Northwestern: Big Ten empty week is after Neb. (filled by Rice, actual Bye before Illinois)
- OSU: After Illinois (somewhat of a rival) and before Wisconsin
- PSU: Before Neb.
- Purdue: Before ND (ND filled Purdue's Big Ten bye, technically after ND)
- Wisconsin: After Neb.
You don't see any bye weeks surrounded by Purdue, Indiana, NW, Minn. You do however see a ton of Nebraska, Wisconsin, and OSU, the three teams expected to be the best. In fact, 8 of the 12 have one of those 3 teams directly before or after their bye. The ones that don't: Illinois (who has PSU then Michigan), Michigan (who had MSU, a rival, before), Purdue (who had their bye week to start and played ND, a rival, during that time), and Iowa (who had it before PSU, a budding rivalry).
October 28th, 2011 at 12:16 PM ^
You're likely right. Thanks for explaining. I have a bit of doubt, though. Given your theory that it relates to the game either before or after, that's two relevant games for each bye. So if there are (as you suggest) three teams that are expected to be the best, out of 11 potential foes, there's almost a 6/11 chance that one of those teams will be the opponent either immediately before or after. In many cases they are not there, however (e.g., Michigan plays none of the three before or after the bye). If you add in "rivalries" either before or after, the odds of facing one of the top three or a rival either before or after the bye becomes pretty good (assuming 1 rival each, which under-counts, it would be almost 8/11.) Even with that, a number would not qualify as playing either the top-3 or the sole rival, e.g., IL, Iowa, UM, Purdue). Statistically, can you be certain that this is a B1G strategy and not just random?
October 28th, 2011 at 12:45 PM ^
makes the sense.
October 28th, 2011 at 1:29 PM ^
You make good points. I just think that it makes sense both from the team and conference perspective that they give teams byes before or after big games. This is what I think is happening in the SEC with Bama and LSU, and I think it tends to happen elsewhere.
Either way, I think the the help a bye week gives a team as far as preperation and health or evened out by rust, over-preperation, and sometimes a lack of urgancy from players. I still think a bye week is a non-factor in games such as this.
October 28th, 2011 at 9:49 AM ^
October 28th, 2011 at 1:31 PM ^
We have talked it up both weeks.
Hoke has a winning record and Dantonio was 0-fer until we played them.
October 28th, 2011 at 7:59 AM ^
This seems like bullshit ESPN statistics "Albert Pujols is batting .560 in Cincinatti the night after the State of the Union address." We're playing Purdue at home. I don't care if Hoke was 0-14 after a bye week, we should win this game.
October 28th, 2011 at 8:51 AM ^
That Pujols stat would most definitely be one of the most bullshit statistics of all time...considering that the SOTU address is usually given in January!
October 28th, 2011 at 8:16 AM ^
A record of 6-4 is pretty darn impressive for Hoke, considering where he was coaching at the time.
I expect his record at Michigan after OPEN weeks will be significantly better than .600.
October 28th, 2011 at 11:15 AM ^
Open weeks normally occur during the conference schedule. This means Ball State would have been playing some MAC team and SDSU would be saying some MWC game. Going 6-4 there is good (I guess, if having a .600 conference record in the MAC/MWC is generally "impressive" over a 10 game period) but it has absolutely no predictive value as to how Hoke will perform against B10 competition at Michigan following an open week.
Plus, the level of competition within the conference following an open week is always rotating. If you randomly get consistently paired up with the best in the conference or the worst in the conference following an open week you will have a worse/better chance of having a higher winning percentage in those weeks, irrespective of the occurrence of the open week. Your comment makes no sense at all.
October 29th, 2011 at 10:21 AM ^
For the love of everything. I never said or have thought his record would have any predictive value.
I think Hoke's going to have a better winning percentage than 60% as Michigan coach. As a result, I think his record after open weeks will be better tha 60%.
Not everything written is underlay with statistical probability and double regression analysis. All I was saying is that I think Hoke is going to win at a high level at MICH.
October 28th, 2011 at 11:20 AM ^
I expect his record at Michigan after OPEN weeks will be significantly better than .600.
Keep fighting the good fight jg2112! This country needs you!
October 28th, 2011 at 8:35 AM ^
Small sample size, but I appreciate you looking them up!
October 28th, 2011 at 8:57 AM ^
WE HAVE A BETTER TEAM THAN PURDUE.
October 28th, 2011 at 10:16 AM ^
where is the knowledge's weekly diary before a game? I needs it
October 28th, 2011 at 10:40 AM ^
October 28th, 2011 at 1:49 PM ^
Thanks for pulling that! Do you mind giving us a winning perecentage too? I don't like math.
/s happy friday
October 28th, 2011 at 10:40 AM ^
While the sample size with Hoke is small the overall numbers are interesting. The .480 number for all teams, while close to 50%, is still negative. I would assume that is a large sample size and indicates a slight disadvantage after a bye. It would be interesting to see the details on the Big Ten’s .350 number.
I did a quick search and someone looked into the issue in the SEC. http://www.teamspeedkills.com/2010/4/15/1424019/how-much-do-bye-weeks-matter
They looked at the last ten years and focused on games that were perceived as toss-ups. They found the team coming off the bye week had a .406 win percentage. More analysis would be needed but their does seem to be a slight disadvantage to a bye week. That seems to go against what we would expect.
I also checked the NFL. The win percentage there is .527 – a slight advantage shown over 586 games of data. Questions: If we were able to remove other factors from these games is there really an advantage or disadvantage to bye weeks? Why is it an advantage in the NFL and a disadvantage in college? Is the Big Ten really that bad after a bye week or is it just related to scheduling? If it is that bad, why?
October 28th, 2011 at 11:17 AM ^
Couple thoughts, for what they're worth:
Hoke mentioned in his Presser that sometimes coaches try to install too much during the bye. I think with college players, it may be better to keep momentum going and not combine a lot of cerebral challenges with rust. Better to promote muscle memory than new plays, formations, etc.
With regard to NFL, perhaps the longer season makes a difference? There's a greater advantage, perhaps, to a rest after around 8 games than 6? Also, less tendency to add more wrinkles than would be useful, and the players already have the muscle memory.
No answers, just stabs in the dark.
October 28th, 2011 at 11:35 AM ^
I mentioned this in one of the other threads, but if you want to see the effect of a bye week, you can't just look at the win percentage. Let's consider the idea that there is a 50/50 chance that a conference bye week precedes a team that is better or worse than the team with the bye week. The AD, then has the option to schedule an OOC game during that conference bye week or at the beginning of the season. If the opponent that the bye week precedes is better than his team, I think it stands to reason there is a greater chance than if the opponent was a worse team that he will leave that bye week to give his team a better chance of beating the better team. The other option is to fill that bye week and give a bye week before your first conference opponent or an OOC foe who may or may not be better. If that is the case, you would theoretically see a lower win percentage against those same opponents if there was no bye week. So, either we see no effect or we see a good effect, IMO.
The way to test this out is to "predict" the winner of the game so that you can compare what the winning percent is against that. If the outcome is expected, it's a tie. If the team with the bye week was expected to lose wins, it's a win. If the team with the bye week was expected to win but loses, it's a loss. Then, compare the win/loss ratio and see if you have more wins. There are obviously more quantitative ways to do this, but it would probably be something easy to setup in a program to do for you.
October 28th, 2011 at 11:40 AM ^
...it was in completely different conference, at a totally different place. That being said, I do think Michigan will take care of Purdue. Purdue's wins came against a 2-3 Middle Tennessee, a 2-5 Southeast Missouri State, a 1-6 Minnesota. Yes, they did also pull out a win against Illinois, but I think the Illini are a bit overrated and Purdue got lucky. Purdue has a so-so offense and their D is suspect. Defensively, they do not stack up to MSU, OSU, WI, etc. They simply do not have the top notch recruits. Denard should have a decent game, we should be able to establish the run and our D should not have much of a problem taking care of business.
October 28th, 2011 at 11:40 AM ^
Did anyone notice that we have 2 bye weeks in 2013. We already have a full schedule of 12 opponents so we can't really fill one of those slots. The second one will be more critical than the first though.
Central Michigan
Notre Dame
Akron
@UCONN
BYE
Minnesota
@Penn State
Indiana
BYE
@Little Brother
Nebraska
@Northwestern
@Iowa
OSU
*Big Ten Championship
October 28th, 2011 at 11:44 AM ^
nice!!!