Distilling an Upset, or How the Vols Fell to A&M (but not Mizzou)

Tennessee has a rematch Saturday with Texas A&M; a chance to avenge the Vols' only loss so far in the month of February. Since dropping that game in College Station, UT has beaten Arkansas by 29 and Vanderbilt by 35. But there was also Tuesday's five-point win at Missouri, a game the Tigers led at halftime and still led with as little as 11 minutes left. It was nearly Tennessee's second upset loss of the month, until the Vols took over in the last 10 minutes, building a lead as large as 13 points before finally winning by 5. 

How did Tennessee avoid the upset loss to Mizzou, but not to the Aggies? The easy answer is that A&M is simply a better team than Missouri, and that's true. TAMU is a bubble team with a solid shot at making the field of 68 (if they can pull out of the nosedive they've been in since beating UT), while Missouri has a solid shot at going 0-fer in the SEC. But there's more to it.

Every team has something they do well, and they will probably do that thing at least passably well regardless of the competition. Texas A&M, for example, is the best rebounding team in America. A&M gets an offensive rebound over 43% of the time. That's huge--almost half the time they miss a shot, they get another chance. You know that when you play the Aggies, they're going to get a lot of offensive glass and some second chance points. That's just going to happen. 

What you cannot let happen, though, is allowing teams to be good at what they're bad at. That's what happened when the Vols last visited College Station. As good as the Aggies are on the offensive glass, they are equally awful from behind the arc. TAMU is bottom 10 nationally in 3-point %. That's easily last in the SEC; the only other league team in the bottom 75 is Vandy. However A&M started out 8-12 from 3 against the Vols and built a lead they'd never relinquish. Looking at a win probability chart, you can see it happen:

1. Tennessee begins the game as 3.5-point favorites with a 67.7% win probability. 
2. That quickly changes as A&M hits a three on their first shot and adds two more treys to open up an 11-8 lead in the first four minutes of the game. Tennessee's win prob. drops to 56.8%.
3. The Aggies blow the game open in the middle part of the 1st half, hitting three more triples in a 12-1 run, dropping UT's win prob. from 50% to less than 25%. Like the little mountain climber guy from The Price is Right, but in the wrong direction. 
4. Even after A&M's shooting cools a bit, UT only manages one slight blip of positivity, when the lead is cut to 12 points by a ZZ 3-pointer. This was around the time Buzz Williams goes into full clown mode, galloping to the panhandle of the midcourt logo to call a timeout. 

Tennessee allowed A&M to be great at something they are normally very bad at. Going into the game, I'm sure Tennessee would have loved A&M to shoot 12 first-half threes, because they would typically only make about three of them. And credit to the Aggies, they shot very well. But defensively, you've got to close out, and you've got to fight through screens or switch off of them. UT didn't do those things effectively against TAMU. 

In a similar way, Tennessee nearly catastrophically lost to SEC-winless Missouri on Tuesday, again because they were letting the underdog do things they're bad at well. Two of the "Four Factors" of basketball are free throw rate (FTA/FGA) and offensive rebound %. Mizzou is bad at both--like bottom-third-nationally bad in each of those categories. On Tuesday, however, Missouri played over their heads in both stats, nearly knocking off the Vols in the process. 

1. Tennessee starts this game as 11.5-point favorites with a 90.5% win probability. Missouri comes into the game with a 29% free throw rate and 26% on offensive rebounds. 
2. After trending the wrong way for most of the first half, Tennessee's win probability falls by over 10 points (to 81.1%) by halftime. At this point, Mizzou has a 52% FTRt and 35% OR%, as well as a 2-point lead. 
3. With about 15:00 left in the game, win probability is in a tailspin, hitting a game-low 65.1%. ESPN's probability calculator (Fig. A) has the Vols' chances at this point even lower, falling below the 50% mark. Missouri has a 37% FTRt and 40% OR% at this point. They are doing very well at what they usually do very bad. If this continues, UT will become the Tigers' first SEC win of the season. 
4. At about the 10:00 mark Tennessee finally breaks through. A 14-2 Vol run cracks the game open and boosts the Vols win prob. to 99.3%. What changes during this time? Missouri only gathers two offensive rebounds and makes one trip to the line. In other words, they go back to being "[expletive] terrible" (Fig. B) at drawing fouls, despite how good they've been at it the previous 31 minutes. 

So yeah, A&M is a better team than Missouri. But the biggest difference in these two games is that Tennessee cut Missouri off from doing what was winning them the game. That didn't happen against  the Aggies. The lesson from these two games is that teams are going to do the thing they're good at. You have to be prepared for that inevitability. What you cannot allow is for teams to also do the things they're bad at well. It's the formula for an upset. Tennessee gets a chance to atone for those mistakes against Texas A&M in Knoxville on Saturday night.