Evaluating Strength of Schedules
I don't have a date for the original date of publication. However, as you will read, it was an attempt to break down how to compare teams who had played easier/harder schedules. Reading back on it many years later, I think the methodology is sound, if rudimentary.
On the message board, we were discussing the relative power of the Big Ten in terms of overall record the last six years.
I posted some numbers to show how Wisconsin compares to some of the other schools, in this case Michigan and Michigan State. Michigan, to see how far or close Wisconsin is to becoming one of the elite teams in the conference, and Michigan State to show whether or not Wisconsin was a step above them in the league hierarchy.
Overall, here are the cumulative records over the last six years in conference play, which is generally the best measure when comparing teams head to head.
Michigan: 35-13 (.729%)
Wisconsin: 28-17-3 (.622 %)
Michigan State: 25-22-1 (.532 %)
(NOTE: For discussions sake, Wisconsin lost to MSU in ‘95, but it is in the books as a win due to MSU forfeiting the game. In this argument, I have left it as a MSU win.)
At this point, a number of posters added that these record are NOT directly comparable, because Michigan plays the “Big Three” every year, while Wisconsin NEVER plays all of the “Big Three” in any year, with MSU somewhere in between, thereby skewing the records. In particular, some MSU defenders claimed that MSU has played a much tougher CONFERENCE schedule over the years (again...due to Wisconsin “getting lucky” with who they miss in a particular season).
So, let’s find out if this explanation has any validity or not. Here are the raw numbers of opponents winning percentage in Big Ten games.
Michigan opponents: 183-192-9 (.488%)
Wisconsin opponents: 178-200-7 (.471%)
Michigan State opponents: 174-201-9 (..464%)
CONCLUSION: Michigan HAS in fact played the toughest schedule of these teams over the last six years, but in fact Wisconsin has played a TOUGHER schedule than MSU. However, the edge is very slight. Michigan, playing the so called cream of the crop every year has only a 5 game in 6 year (opponents wins) advantage over Wisconsin.
Now, is the issue dead? No it is not. Let’s say a team goes 8-0 in conference play as the ‘97 Wolverines did. Compare them to say a team that goes 0-8 like Northwestern last year. Michigan’s opponents are going to be at an immediate 8 game disadvantage, because they were 0-8 against Michigan. Northwestern opponents meanwhile, are going to look even tougher because they went 8-0 against a hapless Northwestern team. Catch my drift?
So, what I did is I took out all games that the team in question themselves played in. So, Michigan is ONLY evaluated on how its opponents did in game OTHER THAN against Michigan. Otherwise, a good team is always going to be penalized more than a bad team, because they are going to worsen their opponents record by beating them silly.
So, here are the revised records of each team's opponents in games not including the team in question.
Michigan opponents: 170-158-9 (.518%)
Wisconsin opponents: 161-172-4 (.483%)
Michigan State opponents: 152-176-8 (.463%)
CONCLUSION: The Michigan opponents take a legit step up when you take out all their losses to Michigan themselves. They definitely have played a tougher conference schedule than either MSU or WIS. Now, how significant is that? More to come. MSU on the other hand, continues to have the weakest conference schedule, falling 5% behind Michigan and 2 % behind Wisconsin. Again, a factor, but not a big one. Now, how could this happen (such close numbers) if Michigan is always playing such a tough schedule and MSU seemingly plays the “Big Three” more than Wisconsin does? The answer is pretty easy actually.
First, in Michigan’s case, they are ONLY playing TWO of the Big Three every year at most....since obviously they can’t play themselves. This matches up exactly with Wisconsin in recent history and is actually lower than MSU, who in some years has played 3 games against the Big Three. Then how does MSU fall so far back? Well first, they don’t play the Big Three every season. They have missed OSU in ‘95 and ‘96 for instance, leaving them with only 2 games against the Big Three in these years.
Secondly, AND MOST IMPORANTANTLY, the Big Three are not always the best teams in the conference. Michigan, for one, spent about 4 straight years at 5-3, finishing in the 3, 4, or 5 spots in the conference. In some years, Wisconsin is one of the Big Three. In others, it has been NW. So, when you say MSU had to play Michigan every year, in some years, that was actually a break if they DIDN’T have to play say Northwestern. As an example, MSU missed out on Illinois in ‘93 and ‘94, both bowl teams . They missed both OSU and and undefeated NW team in ‘95. In ‘96, they missed OSU again, AND a 7-1 NW team. IN ‘97, they didn’t have to play bowl bound Iowa, nor bowl bound Wisconsin. You get the picture. If anything, folks should be crying that the SPARTANS have had the easy schedules, missing one and usually two bowl teams ALMOST EVERY SEASON SINCE ‘93. This is why the numbers show what they do.
So, to sum up, of the three teams mentioned here, Michigan has the best record despite playing the toughest conference schedule. They certainly have earned their place at the top of the conference. Wisconsin, despite playing a HARDER conference schedule than MSU, has a better W/L record, proving at least in my book that they deserve consideration as being a step up from MSU. MSU played the easiest schedule and had the worst record. What more can you say?
Next up......non-conference schedules.
I posted some numbers to show how Wisconsin compares to some of the other schools, in this case Michigan and Michigan State. Michigan, to see how far or close Wisconsin is to becoming one of the elite teams in the conference, and Michigan State to show whether or not Wisconsin was a step above them in the league hierarchy.
Overall, here are the cumulative records over the last six years in conference play, which is generally the best measure when comparing teams head to head.
Michigan: 35-13 (.729%)
Wisconsin: 28-17-3 (.622 %)
Michigan State: 25-22-1 (.532 %)
(NOTE: For discussions sake, Wisconsin lost to MSU in ‘95, but it is in the books as a win due to MSU forfeiting the game. In this argument, I have left it as a MSU win.)
At this point, a number of posters added that these record are NOT directly comparable, because Michigan plays the “Big Three” every year, while Wisconsin NEVER plays all of the “Big Three” in any year, with MSU somewhere in between, thereby skewing the records. In particular, some MSU defenders claimed that MSU has played a much tougher CONFERENCE schedule over the years (again...due to Wisconsin “getting lucky” with who they miss in a particular season).
So, let’s find out if this explanation has any validity or not. Here are the raw numbers of opponents winning percentage in Big Ten games.
Michigan opponents: 183-192-9 (.488%)
Wisconsin opponents: 178-200-7 (.471%)
Michigan State opponents: 174-201-9 (..464%)
CONCLUSION: Michigan HAS in fact played the toughest schedule of these teams over the last six years, but in fact Wisconsin has played a TOUGHER schedule than MSU. However, the edge is very slight. Michigan, playing the so called cream of the crop every year has only a 5 game in 6 year (opponents wins) advantage over Wisconsin.
Now, is the issue dead? No it is not. Let’s say a team goes 8-0 in conference play as the ‘97 Wolverines did. Compare them to say a team that goes 0-8 like Northwestern last year. Michigan’s opponents are going to be at an immediate 8 game disadvantage, because they were 0-8 against Michigan. Northwestern opponents meanwhile, are going to look even tougher because they went 8-0 against a hapless Northwestern team. Catch my drift?
So, what I did is I took out all games that the team in question themselves played in. So, Michigan is ONLY evaluated on how its opponents did in game OTHER THAN against Michigan. Otherwise, a good team is always going to be penalized more than a bad team, because they are going to worsen their opponents record by beating them silly.
So, here are the revised records of each team's opponents in games not including the team in question.
Michigan opponents: 170-158-9 (.518%)
Wisconsin opponents: 161-172-4 (.483%)
Michigan State opponents: 152-176-8 (.463%)
CONCLUSION: The Michigan opponents take a legit step up when you take out all their losses to Michigan themselves. They definitely have played a tougher conference schedule than either MSU or WIS. Now, how significant is that? More to come. MSU on the other hand, continues to have the weakest conference schedule, falling 5% behind Michigan and 2 % behind Wisconsin. Again, a factor, but not a big one. Now, how could this happen (such close numbers) if Michigan is always playing such a tough schedule and MSU seemingly plays the “Big Three” more than Wisconsin does? The answer is pretty easy actually.
First, in Michigan’s case, they are ONLY playing TWO of the Big Three every year at most....since obviously they can’t play themselves. This matches up exactly with Wisconsin in recent history and is actually lower than MSU, who in some years has played 3 games against the Big Three. Then how does MSU fall so far back? Well first, they don’t play the Big Three every season. They have missed OSU in ‘95 and ‘96 for instance, leaving them with only 2 games against the Big Three in these years.
Secondly, AND MOST IMPORANTANTLY, the Big Three are not always the best teams in the conference. Michigan, for one, spent about 4 straight years at 5-3, finishing in the 3, 4, or 5 spots in the conference. In some years, Wisconsin is one of the Big Three. In others, it has been NW. So, when you say MSU had to play Michigan every year, in some years, that was actually a break if they DIDN’T have to play say Northwestern. As an example, MSU missed out on Illinois in ‘93 and ‘94, both bowl teams . They missed both OSU and and undefeated NW team in ‘95. In ‘96, they missed OSU again, AND a 7-1 NW team. IN ‘97, they didn’t have to play bowl bound Iowa, nor bowl bound Wisconsin. You get the picture. If anything, folks should be crying that the SPARTANS have had the easy schedules, missing one and usually two bowl teams ALMOST EVERY SEASON SINCE ‘93. This is why the numbers show what they do.
So, to sum up, of the three teams mentioned here, Michigan has the best record despite playing the toughest conference schedule. They certainly have earned their place at the top of the conference. Wisconsin, despite playing a HARDER conference schedule than MSU, has a better W/L record, proving at least in my book that they deserve consideration as being a step up from MSU. MSU played the easiest schedule and had the worst record. What more can you say?
Next up......non-conference schedules.
Part 2
In part one, I posted some numbers that showed that despite playing a tougher Big Ten schedule than Michigan State over the past 6 years, Wisconsin had a higher winning percentage than Michigan State. This was done to refute the Sparties claim that Wisconsin’s success in conference was due in part to their cush schedule. Well, as it turned out, the egg was on the Sparties as Wisconsin’s conference schedule ended up being more difficult.
Well, what about the OVERALL difficulty of schedule being slanted heavily towards Wisconsin? This of course has been a consistent slam on the Wisconsin program, compiling impressive records by playing dregs in the non-conference. MSU supporters also added that due to a “much” tougher NC schedule, that this played an adverse part in the Sparties Big Ten play (more banged up, emotionally drained, etc.). They also threw in the fact that MSU generally beats better teams...that essentially they have a higher ceiling of play.
Well, as always, let’s look at the numbers with these key questions in mind:
1.) Does Wisconsin in fact play an easier non-conference schedule than Michigan and Michigan State?
2.) If so, how much tougher have the MSU and Michigan schedules been and how has Wisconsin done against similar competition?
3.) Is this difference in competition enough to explain the difference between WIS and MSU’s overall record in the last six years and what would MSU’s record be playing Wisconsin’s schedule?
First, how can we evaluate schedules? Unlike conference records, where opponents W/L percentage can be used (since all the data is taken in the same context), winning percentage in non-conference games is trickier since some teams can compile vastly different records depending on the strength of THEIR schedules. In short, an 9-2 SEC team is not equivalent to a 9-2 WAC team. So instead, I am using end of the year power rankings. The particular equation I am using is the Howell computer power rankings. They are very similar to Sagarin’s ratings with a few slight differences. I am using these as they are readily available to me for the last umpteen years worth of seasons.
First, here is the average strength of opponent over the last 6 years:
Wisconsin: 47
Michigan State: 41
Michigan: 39
CONCLUSION: Wisconsin has in fact had the easiest overall schedule in terms of average strength of opponent over the last six years. Michigan has played the toughest schedule of the three, though by a pretty small margin over the Spartans.
This brings us to question #2: what exactly do these numbers mean. Is this a significant difference? How has this difference affected records? What I did next is broke it down into tiers: how did each team do against different levels of teams? Has Wisconsin not played the elite teams as often as Michigan or Michigan State? Is it a problem at the bottom end? Has MSU been the better team but had this fact hidden by the tougher schedule?
Record against Top 10 teams
Michigan: 6-7 (.462 winning percentage)
Wisconsin: 2-8-1 (.200)
Michigan State:1-10 (.091)
CONCLUSION: Wisconsin had played EXACTLY THE SAME NUMBER OF GAMES VS. TOP 10 TEAMS as has MSU over the last 6 years. Wisconsin has also fared slightly better in these games than have the Spartans. Michigan has played 2 more games vs. Top 10 schools over the last 6 years with more success than either WIS or MSU.
Record against teams 11-20
Michigan: 8-4 (.667)
Wisconsin: 6-6 (.500)
Michigan State: 3-12 (.200)
CONCLUSION: Michigan has done pretty well against top level (though not elite) teams over the years. Wisconsin has played the exact same number of teams 11-20 over the last 6 years, and has held their own with a .500 record. MSU however, while playing 3 more of these games has been dreadful in such games, getting pretty much hammered.
Combining these two figures into one stat (Top 20 teams), we get this:
Michigan: 25 games (14-11 record , .560)
Wisconsin: 23 games (8-14-1, .364)
Michigan State: 26 games (4-22, .154)
CONCLUSION: Contrary to popular belief, all three teams have played about the same amount of games vs. Top 20 teams. Michigan clearly has been the best team against upper level teams, Wisconsin has done OK (nothing special), while MSU has gotten drilled.
OK, how about the solid, mid level teams?
Teams ranked 21-40
Wisconsin: 8-4-2 (.667)
Michigan: 12-7 (.632)
Michigan State: 5-6 (.455)
CONCLUSION: Wisconsin actually has the best record of the three vs. the upper middle type teams. Michigan has played the most games vs. such teams, with Wisconsin #2. Michigan State has played the fewest games against upper middle teams and still have not reached .500 against any group of teams we have looked at yet.
Combined totals (records vs. teams 1-40)
Michigan: 44 games, 26-18, .591
Wisconsin: 37 games, 16-18-3, .471
Michigan State: 37 games, 9-28, .243
CONCLUSION: Well Spartan fans...I hate to say it but your theory is totally shot. Wisconsin has played EXACTLY the same number of games vs. top 40 teams, with much more success.
Teams ranked 41-60 (lower middle teams)
Michigan: 10-0 (1.000)
Wisconsin: 5-1-1 (.833)
Michigan State: 7-4-1 (.636)
CONCLUSION: Hey, MSU finally got over .500 against somebody, though still way behind Michigan’s impressive record and also behind Wisconsin’s. MSU played 11 such games while Wisconsin played 7, a slight difference, though considering what teams we are talking about is hardly a big deal. (but we will take it into account later)
Teams ranked 61+
Michigan: 19-1 (.950 )-FYI-the one loss was against Purdue
Wisconsin: 27-2 (.931)
Michigan State: 20-2 (.909)
CONCLUSION: MSU finally dominates on somebody, though they are STILL behind Wisconsin and Michigan. Wisconsin has played the most games against such teams with 29, as opposed to 22 from MSU and 20 from Michigan. Take the 5 games from the previous set of data to MSU’s “credit” and you are left with 2 “extra” games, basically saying that Wisconsin played two games against cupcakes when MSU wasn’t playing anybody (let’s call those the two Hawaii games).
Wisconsin beats MSU in EVERY SINGLE BREAKDOWN, regardless of who the opponent was. It simply doesn’t matter, Wisconsin has been a better team against great, good, average, and poor teams. By the same token, Wisconsin only beat Michigan in one category. Michigan clearly is the best of the group with Wisconsin a very clear second. Not only did Wisconsin play better against good teams, but they have also played roughly the same number of games against good teams as has MSU. They have played the same number of games against Top 40 teams, and against Top 10 teams. MSU has a very slight edge (3 games) against teams 11-20 but Wisconsin makes it up on teams 21-40. Both teams have however played a very slightly less difficult schedule than Michigan in the same time frame. However, this edge essentially amounts to one game a year.
So, basically both MSU AND WIS are playing a bottom feeder while Michigan is playing one Top 40 (note...not top 10 or even top 20) team. While this COULD be a difference, at most it is one win a year every other year for Wisconsin (based on their winning percentage against each group) and about 2 wins every 3 years for the Sparties.
So, if all this info basically matches up between MSU and WIS, how did MSU’s average opponent rank higher than Wisconsin’s (way back at the beginning)? Well, the answer is easy actually, and it comes down to that last grouping of teams 61+. Whereas Wisconsin has played some really bad stinkers (ranking in the 100’s) at times, MSU generally has ALSO played a bottom feeder, though one not quite as bad (usually in the 80’s). This “playing the better bottom feeder” has basically served to give MSU the edge in terms of average strength of opponent. Since both teams clean house against these scrubs, it really hasn’t mattered at all in the grand scheme of things (ie. records.)
Now, for the final question: what would Michigan’s and Michigan State’s record been if playing Wisconsin’s schedule? Let's try to normalize the schedules so that each team is playing the same number of game within the same tier.
Let's begin with Michigan State: Both Wisconsin and Michigan State played 11 games against top 10 teams, so nothing needs to be adjusted there.
MSU’S HYPOTHETICAL RECORD: 1-10
Next, against teams 11-20, MSU played three more games than Wisconsin. Let’s take out those 3 games and put them into the “mystery games” (meaning game of unknown result) category. Those 3 games all likely were losses (based on actual winning percentage), so we subtract them from MSU's loss column, leaving MSU’s HYPOTHETICAL RECORD at 4-19 (with 3 “mystery games”).
Next are games 21-40. Wisconsin played 3 MORE of these games than MSU. So, to even things out, put the 3 “mystery games” that MSU had banked here. Based on winning percentage, 1 of these games would now become a win, while the other 2 would remain losses.
MSU’S HYPOTHETICAL RECORD: 10-27.
In games 41-60, MSU played 5 more games vs. these teams vs. Wisconsin. So, put these 5 games into the “mystery games” category. Of these 5 games, 3 would be wins and 2 would be losses so we remove those from the record.
MSU’S HYPOTHETICAL RECORD: 14-29-1
In games against teams ranked 61+, WIS has played 7 more games than MSU. First, those 5 banked “mystery games” are added to this category. Throw in the 2 games that Wisconsin played that MSU didn’t, and you get a total of 7 games added against lightweights. 6 would be wins, 1 would be a loss for MSU.
MSU’S HYPOTHETICAL RECORD: 40-32-1
So, compare this to MSU’s actual record of 36-34-1 and you see that MSU would in fact have been a bit better, in fact gaining 4 games over 6 years. Based on part 1, we can assume these would be non-conference games.
Wisconsin’s record over that time: 48-21-4.
CONCLUSION: MSU would indeed have a better record if they had played Wisconsin’s schedule. However, the net difference would NOT be enough to make up the difference in their records. MSU would still be 8 games behind, which comes out to a bit more than 1 win per year. The gap would simply be a bit tighter.
Now, let’s try this with Michigan.
Michigan has played 13 top 10 games. Wisconsin has played 11. Michigan gets 2 to the “mystery games” column. 1 would be a win. 1 would be a loss.
MICHIGAN’S HYPOTHETICAL RECORD: 5-6
Each team has played 12 games against the 11-20 teams, so these results stay the same:
MICHIGAN’S HYPOTHETICAL RECORD: 13-10.
Against teams 21-40, Michigan played 5 more of these games than Wisconsin. So, they get to add these to the “mystery game” column, for a total of 7. Of these games, 3 would be wins, 2 would be losses.
MICHIGAN’S HYPOTHETICAL RECORD: 22-15.
In the 41-60 range of teams, Michigan played 3 more than Wisconsin. Again, add them to the “mystery game” column, for a total of 10. All would be wins.
MICHIGAN’S HYPOTHETICAL RECORD: 29-15
Finally, we have the games against the scrubs, teams 61+. Wisconsin has played 9 more games than Michigan against these teams. So, take 9 of Michigan’s “mystery games” and put them for Michigan here (against the cupcakes). All would be wins.
MICHIGAN’S HYPOTHETICAL RECORD: 57-16.
The final game would be thrown out (since we are matching Wisconsin’s schedule).
Michigan’s actual record in this time was 55-19. So, they would gain an additional 2 wins in 6 years while losing 1 loss. Why the rather limited gain? Easy again. It is because in many of these games, Michigan WON ANYWAY!!! Moving a game against a #38 team (probably a win) and making it a game against a patsy UNLV team at #110, still results in a win, so no change in record. Basically speaking, Wisconsin HAS had an easier schedule, but the statistical impact is very small.
It does serve to show exactly how each team measures in relation to each other since we have now controlled for strength of schedule.
PROJECTED RECORDS
Michigan: 57-16 (.781)
Wisconsin: 48-21-4 (.696)
Michigan State: 40-32-1 (.556)
I have always said that Wisconsin has been creeping up on Michigan in terms of overall level. While they are improving, they certainly have a ways to go (.085 % points) and I think my comment in that regard may have been premature. However, I am even more steadfast in my conclusion that Wisconsin is a legit step above MSU (.140 % points).
Hope you enjoyed this....perhaps the longest post in sports board history!
Well, what about the OVERALL difficulty of schedule being slanted heavily towards Wisconsin? This of course has been a consistent slam on the Wisconsin program, compiling impressive records by playing dregs in the non-conference. MSU supporters also added that due to a “much” tougher NC schedule, that this played an adverse part in the Sparties Big Ten play (more banged up, emotionally drained, etc.). They also threw in the fact that MSU generally beats better teams...that essentially they have a higher ceiling of play.
Well, as always, let’s look at the numbers with these key questions in mind:
1.) Does Wisconsin in fact play an easier non-conference schedule than Michigan and Michigan State?
2.) If so, how much tougher have the MSU and Michigan schedules been and how has Wisconsin done against similar competition?
3.) Is this difference in competition enough to explain the difference between WIS and MSU’s overall record in the last six years and what would MSU’s record be playing Wisconsin’s schedule?
First, how can we evaluate schedules? Unlike conference records, where opponents W/L percentage can be used (since all the data is taken in the same context), winning percentage in non-conference games is trickier since some teams can compile vastly different records depending on the strength of THEIR schedules. In short, an 9-2 SEC team is not equivalent to a 9-2 WAC team. So instead, I am using end of the year power rankings. The particular equation I am using is the Howell computer power rankings. They are very similar to Sagarin’s ratings with a few slight differences. I am using these as they are readily available to me for the last umpteen years worth of seasons.
First, here is the average strength of opponent over the last 6 years:
Wisconsin: 47
Michigan State: 41
Michigan: 39
CONCLUSION: Wisconsin has in fact had the easiest overall schedule in terms of average strength of opponent over the last six years. Michigan has played the toughest schedule of the three, though by a pretty small margin over the Spartans.
This brings us to question #2: what exactly do these numbers mean. Is this a significant difference? How has this difference affected records? What I did next is broke it down into tiers: how did each team do against different levels of teams? Has Wisconsin not played the elite teams as often as Michigan or Michigan State? Is it a problem at the bottom end? Has MSU been the better team but had this fact hidden by the tougher schedule?
Record against Top 10 teams
Michigan: 6-7 (.462 winning percentage)
Wisconsin: 2-8-1 (.200)
Michigan State:1-10 (.091)
CONCLUSION: Wisconsin had played EXACTLY THE SAME NUMBER OF GAMES VS. TOP 10 TEAMS as has MSU over the last 6 years. Wisconsin has also fared slightly better in these games than have the Spartans. Michigan has played 2 more games vs. Top 10 schools over the last 6 years with more success than either WIS or MSU.
Record against teams 11-20
Michigan: 8-4 (.667)
Wisconsin: 6-6 (.500)
Michigan State: 3-12 (.200)
CONCLUSION: Michigan has done pretty well against top level (though not elite) teams over the years. Wisconsin has played the exact same number of teams 11-20 over the last 6 years, and has held their own with a .500 record. MSU however, while playing 3 more of these games has been dreadful in such games, getting pretty much hammered.
Combining these two figures into one stat (Top 20 teams), we get this:
Michigan: 25 games (14-11 record , .560)
Wisconsin: 23 games (8-14-1, .364)
Michigan State: 26 games (4-22, .154)
CONCLUSION: Contrary to popular belief, all three teams have played about the same amount of games vs. Top 20 teams. Michigan clearly has been the best team against upper level teams, Wisconsin has done OK (nothing special), while MSU has gotten drilled.
OK, how about the solid, mid level teams?
Teams ranked 21-40
Wisconsin: 8-4-2 (.667)
Michigan: 12-7 (.632)
Michigan State: 5-6 (.455)
CONCLUSION: Wisconsin actually has the best record of the three vs. the upper middle type teams. Michigan has played the most games vs. such teams, with Wisconsin #2. Michigan State has played the fewest games against upper middle teams and still have not reached .500 against any group of teams we have looked at yet.
Combined totals (records vs. teams 1-40)
Michigan: 44 games, 26-18, .591
Wisconsin: 37 games, 16-18-3, .471
Michigan State: 37 games, 9-28, .243
CONCLUSION: Well Spartan fans...I hate to say it but your theory is totally shot. Wisconsin has played EXACTLY the same number of games vs. top 40 teams, with much more success.
Teams ranked 41-60 (lower middle teams)
Michigan: 10-0 (1.000)
Wisconsin: 5-1-1 (.833)
Michigan State: 7-4-1 (.636)
CONCLUSION: Hey, MSU finally got over .500 against somebody, though still way behind Michigan’s impressive record and also behind Wisconsin’s. MSU played 11 such games while Wisconsin played 7, a slight difference, though considering what teams we are talking about is hardly a big deal. (but we will take it into account later)
Teams ranked 61+
Michigan: 19-1 (.950 )-FYI-the one loss was against Purdue
Wisconsin: 27-2 (.931)
Michigan State: 20-2 (.909)
CONCLUSION: MSU finally dominates on somebody, though they are STILL behind Wisconsin and Michigan. Wisconsin has played the most games against such teams with 29, as opposed to 22 from MSU and 20 from Michigan. Take the 5 games from the previous set of data to MSU’s “credit” and you are left with 2 “extra” games, basically saying that Wisconsin played two games against cupcakes when MSU wasn’t playing anybody (let’s call those the two Hawaii games).
Wisconsin beats MSU in EVERY SINGLE BREAKDOWN, regardless of who the opponent was. It simply doesn’t matter, Wisconsin has been a better team against great, good, average, and poor teams. By the same token, Wisconsin only beat Michigan in one category. Michigan clearly is the best of the group with Wisconsin a very clear second. Not only did Wisconsin play better against good teams, but they have also played roughly the same number of games against good teams as has MSU. They have played the same number of games against Top 40 teams, and against Top 10 teams. MSU has a very slight edge (3 games) against teams 11-20 but Wisconsin makes it up on teams 21-40. Both teams have however played a very slightly less difficult schedule than Michigan in the same time frame. However, this edge essentially amounts to one game a year.
So, basically both MSU AND WIS are playing a bottom feeder while Michigan is playing one Top 40 (note...not top 10 or even top 20) team. While this COULD be a difference, at most it is one win a year every other year for Wisconsin (based on their winning percentage against each group) and about 2 wins every 3 years for the Sparties.
So, if all this info basically matches up between MSU and WIS, how did MSU’s average opponent rank higher than Wisconsin’s (way back at the beginning)? Well, the answer is easy actually, and it comes down to that last grouping of teams 61+. Whereas Wisconsin has played some really bad stinkers (ranking in the 100’s) at times, MSU generally has ALSO played a bottom feeder, though one not quite as bad (usually in the 80’s). This “playing the better bottom feeder” has basically served to give MSU the edge in terms of average strength of opponent. Since both teams clean house against these scrubs, it really hasn’t mattered at all in the grand scheme of things (ie. records.)
Now, for the final question: what would Michigan’s and Michigan State’s record been if playing Wisconsin’s schedule? Let's try to normalize the schedules so that each team is playing the same number of game within the same tier.
Let's begin with Michigan State: Both Wisconsin and Michigan State played 11 games against top 10 teams, so nothing needs to be adjusted there.
MSU’S HYPOTHETICAL RECORD: 1-10
Next, against teams 11-20, MSU played three more games than Wisconsin. Let’s take out those 3 games and put them into the “mystery games” (meaning game of unknown result) category. Those 3 games all likely were losses (based on actual winning percentage), so we subtract them from MSU's loss column, leaving MSU’s HYPOTHETICAL RECORD at 4-19 (with 3 “mystery games”).
Next are games 21-40. Wisconsin played 3 MORE of these games than MSU. So, to even things out, put the 3 “mystery games” that MSU had banked here. Based on winning percentage, 1 of these games would now become a win, while the other 2 would remain losses.
MSU’S HYPOTHETICAL RECORD: 10-27.
In games 41-60, MSU played 5 more games vs. these teams vs. Wisconsin. So, put these 5 games into the “mystery games” category. Of these 5 games, 3 would be wins and 2 would be losses so we remove those from the record.
MSU’S HYPOTHETICAL RECORD: 14-29-1
In games against teams ranked 61+, WIS has played 7 more games than MSU. First, those 5 banked “mystery games” are added to this category. Throw in the 2 games that Wisconsin played that MSU didn’t, and you get a total of 7 games added against lightweights. 6 would be wins, 1 would be a loss for MSU.
MSU’S HYPOTHETICAL RECORD: 40-32-1
So, compare this to MSU’s actual record of 36-34-1 and you see that MSU would in fact have been a bit better, in fact gaining 4 games over 6 years. Based on part 1, we can assume these would be non-conference games.
Wisconsin’s record over that time: 48-21-4.
CONCLUSION: MSU would indeed have a better record if they had played Wisconsin’s schedule. However, the net difference would NOT be enough to make up the difference in their records. MSU would still be 8 games behind, which comes out to a bit more than 1 win per year. The gap would simply be a bit tighter.
Now, let’s try this with Michigan.
Michigan has played 13 top 10 games. Wisconsin has played 11. Michigan gets 2 to the “mystery games” column. 1 would be a win. 1 would be a loss.
MICHIGAN’S HYPOTHETICAL RECORD: 5-6
Each team has played 12 games against the 11-20 teams, so these results stay the same:
MICHIGAN’S HYPOTHETICAL RECORD: 13-10.
Against teams 21-40, Michigan played 5 more of these games than Wisconsin. So, they get to add these to the “mystery game” column, for a total of 7. Of these games, 3 would be wins, 2 would be losses.
MICHIGAN’S HYPOTHETICAL RECORD: 22-15.
In the 41-60 range of teams, Michigan played 3 more than Wisconsin. Again, add them to the “mystery game” column, for a total of 10. All would be wins.
MICHIGAN’S HYPOTHETICAL RECORD: 29-15
Finally, we have the games against the scrubs, teams 61+. Wisconsin has played 9 more games than Michigan against these teams. So, take 9 of Michigan’s “mystery games” and put them for Michigan here (against the cupcakes). All would be wins.
MICHIGAN’S HYPOTHETICAL RECORD: 57-16.
The final game would be thrown out (since we are matching Wisconsin’s schedule).
Michigan’s actual record in this time was 55-19. So, they would gain an additional 2 wins in 6 years while losing 1 loss. Why the rather limited gain? Easy again. It is because in many of these games, Michigan WON ANYWAY!!! Moving a game against a #38 team (probably a win) and making it a game against a patsy UNLV team at #110, still results in a win, so no change in record. Basically speaking, Wisconsin HAS had an easier schedule, but the statistical impact is very small.
It does serve to show exactly how each team measures in relation to each other since we have now controlled for strength of schedule.
PROJECTED RECORDS
Michigan: 57-16 (.781)
Wisconsin: 48-21-4 (.696)
Michigan State: 40-32-1 (.556)
I have always said that Wisconsin has been creeping up on Michigan in terms of overall level. While they are improving, they certainly have a ways to go (.085 % points) and I think my comment in that regard may have been premature. However, I am even more steadfast in my conclusion that Wisconsin is a legit step above MSU (.140 % points).
Hope you enjoyed this....perhaps the longest post in sports board history!
2022 Epilogue
Records since 1993:
Wisconsin: 260-106-4 (.708)
Michigan: 245-114 (.682)
Michigan State: 206-150-1 (.578)
No controls for schedule, but just a follow up on my conjecture about whether Wisconsin could catch Michigan.
Wisconsin: 260-106-4 (.708)
Michigan: 245-114 (.682)
Michigan State: 206-150-1 (.578)
No controls for schedule, but just a follow up on my conjecture about whether Wisconsin could catch Michigan.