Like my previous post said, I was in Vegas over the (long) weekend for a kind of reunion with grad school friends. It’s always a great time, and this trip was no different. And since I was there, I was able to practice a bit of what I preach in terms of putting down some bets. Here’s how things went.
My memory about the model had been that it traditionally does pretty poorly on wild card weekend. Even in seasons where it did well, and even in seasons that were used to fit the model (meaning that the model will do better than expected because they aren’t real predictions), the wild card round just didn’t match up. But after I got back, I looked at how Luigi did the past two years (which is as long as Luigi has been around). He actually went 3-1 and 2-2 against the spread, but is only 4-4 in picking the correct outright winner. So it isn’t like that’s super great, but with only 8 games to go from it seemed ok. So maybe I should have felt a bit more confidence in Luigi.
On the other hand, Luigi had a pretty off year this season, and I also had two other models to go with. Yoshi 2 wasn’t so hot either, but Yoshi 1 had a decent season. So when I decided what bets to make, I used a combination of Luigi, Yoshi 1, and my own feelings about if I felt like risking anything. With the numbers I had available, I decided that Houston-Bengals was a little too close to call even though both models had the Bengals covering by about a point (Yoshi 2 had it just within the ‘magic’ 1 point stay-away boundary). Luigi had Yoshi 1 disagreed about Vikings-Packers, with Luigi saying that the Vikings should cover while Yoshi 1 said that the Packers should cover. That was before it was announced that Ponder was out, although Yoshi 1 still had the Packers covering the -9.5 I saw closer to game time. The two places that all the models agreed was with the Sunday games, where both the Colts and Redskins were supposed to cover. They all also agreed that the underdogs (Vikings, Bengals, Colts, and Redskins) were getting good enough odds to take on the moneyline. I went a bit more with my gut on those, and I didn’t want to have too much money riding on any one outcome, so in the end I put a little bit on the Vikings to win outright (hoping for AP to go crazy and to get a repeat of the week 17 game) and a bit more on the Colts and Redskins each to cover.
Obviously we know what happened, and I knew enough from catching bits of the games while walking around casinos. The Redskins had a great start and were even in range for a backdoor push (occasionally, betting sounds dirty) but simply couldn’t do anything on offense once RGIII was (further) injured. The Vikings never got anywhere, and it’s really unclear if having Ponder would have helped. Peterson ran as often as he did for his first 200 yard game against the Packers, but never broke a big run and wasn’t quite as efficient as he was in his second 200 yard game, where he didn’t have any runs over 28 yards but still averaged nearly 6 yards a pop. And the Colts were simply not especially good against the Ravens. Luck threw the ball over 50 times yet didn’t break 300 yards; Flacco had 6 fewer yards on 31 fewer passes. They couldn’t stop the run or the deep ball, as five different Ravens had catches of over 20 yards. They couldn’t get to Flacco, with only 1 sack and 2 QB hits in the boxscore. It doesn’t help that the Colts settled for a field goal from the Raven’s 8 and missed another one.
So in the end, I put down three bets and none came through. It was a poor weekend overall as both Luigi and Yoshi 2 were 0-4 against the spread; Yoshi 1 was 1-3 (getting the Packers). None of the underdogs won, either. But don’t feel too badly for me; I had a good run at the craps table with some friends and came out ahead for the weekend.
Now let me point something out. I have a pretty steady stream of football posts come out during the season (and occasionally outside it), but only two have garnered any kind of response. One was this article on the Colts, describing how they have the numbers of a poor team but have simply had an outrageously easy schedule, and that’s how they made the playoffs and won so many games. Reaction was kind of split there, with some Colts fans telling me I was right and some people telling me I was obviously crazy. One comment at the end of the season (the article is from back in week 10) even said I was wrong, even though pretty much everything I said turned out to be true (the Colts beat the additional easy teams they played, they had a tough game with the Lions, the Patriots handed them their lunch, they made the playoffs and played one of the Patriots, Ravens, Steelers, or Broncos). The only thing I missed was that they ended up splitting with the Texans instead of losing both games. Now that the Colts have officially lost, it would be easy to say “I told you so”, but that isn’t the point.
The second article, and the reason I can’t say “I told you so”, is this one on KC Joyner and the Giants. Joyner wrote an article where he picked silly stats to say that the Giants should be a favorite in the playoffs last year. The stats were silly because they favored other teams the Giants would have to play as much as they favored the Giants; additionally, virtually no team that has to play in the wild card round should be called a favorite simply because they have that one extra chance to lose. The four teams that get a bye are the favorites, and that’s pretty much that. I caught a lot of flack because the Giants did indeed go on to win the Super Bowl. I think (although I know people will disagree) that it should be obvious that the Giants were fortunate to win. They barely got in the playoffs at all, clinching a berth in week 17. They were only 3 point favorites over Atlanta and were 9 point dogs to the Packers, 3 point dogs to the 49ers (whom they beat in overtime following two fumbles on punt returns by the 49ers’ back-up return man), and 3 point dogs to the Patriots. That is to say, at no point after the first weekend were the Giants ever considered favorites by the people who make a living by evaluating football teams. But I get nothing but crummy comments on that post because the Giants managed to pull out a few games.
So what does that Giants post have to do with the Colts? One facet is that Joyner went out on a limb again this year and wrote about how the Colts should be even with the Ravens if not favored (Insider required). He said the Ravens can’t rely on the run (they ran 32 times for over 5 yards a pop), Reggie Wayne would have a big day (he did, but didn’t score), Flacco can’t throw deep (see what I said earlier about five guys catching passes of over 20 yards), and that the Colts can win the turnover game (each team had two turnovers). You can say that underdogs have a chance every year, but they aren’t going to come through all the time. Joyner went with the Giants last year and it worked; he went with the Colts this year and it didn’t. But the main point about these two articles is that what is important is the process. My models take certain stats, weight them in a manner that best helps them predict future outcomes, and makes a prediction. The Colts didn’t have good stats this year, and the Giants didn’t have great stats last year. Joyner picked some stats he likes, maybe combines them in an unclear manner, and says that’s good enough to pick one team to make a run. Hopefully one of those methods sounds better to you, and hopefully it isn’t Joyner’s.
But they are only methods and predictions; predictions don’t have to turn out correctly. Sometimes crazy things happen. Here’s a short list from this year: the Cardinals beat the Patriots, the end of that Seahawks-Packers game, Adrian Peterson ran for over 2000 yards less than a year after he blew out his ACL, Peyton Manning threw for over 4600 yards and 37 touchdowns with a new team after having neck surgery and not playing NFL football for a full year. I don’t think anyone would have said those things were likely before they happened, even if they thought they were possible. Once they do happen though, we tend to get caught up in outcomes far more quickly than should happen. Of course Manning would be Manning; AP is a supernatural beast of some kind. We ignore the fact that sometimes things happen. And so I’m not going to say “I told you so” to Joyner or the Colts, because that’s getting caught up in the outcome. Maybe they really were even with the Ravens and they just had a bad day. But I will say that I like my process, I’m not the only one who thought the Colts were overrated, and I think the best team won last week. Even still, those Colts managed to cost me money.
Speaking of which, divisional round predictions to come tomorrow! Luigi is 3-1 in those games each of the past two years, so hopefully it’ll be a good week.