:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/65605641/1184240583.jpg.0.jpg)
Last season, after 7 games the Indianapolis Colts were sitting at 2 - 5 and I wrote an article titled “The Colts aren’t as bad as you think”. After I posted that, the team went 8 - 1 to finish the season. At the same point this year, the Colts are 5-2 and it’s only fair to ask if this team is as good as their record. So, I thought I would test the 2019 squad using the same analysis I did in 2018 and perhaps shed some light on the rest of the season.
NOTE: I actually wasn’t going to write this piece, but the Indianapolis Star recently wrote an article in a similar vein and I’ll be damned if I’m going to let them steal all of the math glory.
POWER RANKINGS
In Chris Blystone’s recent power rankings article, 5 sources ranked the Colts between 6th and 11th. That’s an average of 8.6th assuming you can wrap your head around a fractional rank.
Seeing as the Colts actual record places them 9th in the league that ranking seems reasonable, but it also seems a little lazy. Are the pundits simply counting wins and not actually measuring team performance?
ODDS
Currently, Sportsline.com has the Colts odds of winning the Superbowl at 25 to 1. There are only 8 other teams with better odds listed. So, the betting community thinks the Colts are the 9th best team, which agrees with the power rankings.
POINT DIFFERENTIALS
Now, we get to the mathy stuff. At the game level, the point differential defines a win or loss. Well, at the season level the same tends to be true. So far this year (before the Thursday night game), the top 2 teams by record are the undefeated New England Patriots and the undefeated San Francisco 49ers. Unsurprisingly, they have the largest season point differentials beating their opponents by an average of 23.6 and 18.6 points respectively. That’s a little scary.
Where are the Colts? So far, Indianapolis has +1 per game point differential. That is +7 on the season and ranks 15th of all teams (worst in the AFC South).
.
| Rank | Team | Wins | Games | Point Differential | Diff/gme |
|---|---|---|---|---|---|
| Rank | Team | Wins | Games | Point Differential | Diff/gme |
| 1 | NE | 8 | 8 | 189 | 23.6 |
| 2 | SF | 7 | 7 | 130 | 18.6 |
| 3 | MIN | 6 | 8 | 79 | 9.9 |
| 4 | DAL | 4 | 7 | 66 | 9.4 |
| 5 | BAL | 5 | 7 | 58 | 8.3 |
| 6 | GB | 7 | 8 | 52 | 6.5 |
| 7 | KC | 5 | 8 | 45 | 5.6 |
| 8 | LA | 5 | 8 | 40 | 5.0 |
| 9 | NO | 7 | 8 | 39 | 4.9 |
| 10 | HOU | 5 | 8 | 24 | 3.0 |
| 11 | BUF | 5 | 7 | 12 | 1.7 |
| 12 | TEN | 4 | 8 | 13 | 1.6 |
| 13 | SEA | 6 | 8 | 12 | 1.5 |
| 14 | JAX | 4 | 8 | 10 | 1.3 |
| 15 | IND | 5 | 7 | 7 | 1.0 |
| 16 | CHI | 3 | 7 | 6 | 0.9 |
| 17 | PIT | 3 | 7 | 5 | 0.7 |
| 18 | PHI | 4 | 8 | 3 | 0.4 |
| 19 | LAC | 3 | 8 | 0 | 0.0 |
| 20 | CAR | 4 | 7 | -5 | -0.7 |
| 21 | DET | 3 | 7 | -6 | -0.9 |
| 22 | TB | 2 | 7 | -16 | -2.3 |
| 23 | DEN | 2 | 8 | -26 | -3.3 |
| 24 | OAK | 3 | 7 | -41 | -5.9 |
| 25 | ARI | 3 | 8 | -53 | -6.6 |
| 26 | CLE | 2 | 7 | -48 | -6.9 |
| 27 | NYG | 2 | 8 | -60 | -7.5 |
| 28 | ATL | 1 | 8 | -85 | -10.6 |
| 29 | CIN | 0 | 8 | -86 | -10.8 |
| 30 | WAS | 1 | 8 | -96 | -12.0 |
| 31 | NYJ | 1 | 7 | -107 | -15.3 |
| 32 | MIA | 0 | 7 | -161 | -23.0 |
Between 2000 - 2018, there have been 110 teams that started 5-2. All but 5 of them had point differentials higher than +7.
PYTHAGOREAN WINS
Football Outsiders takes point differentials and performs an Adjusted Pythagorean Win Expectation, which estimates the number of wins a team “should have” given their cumulative points for and against. Historically, that measure has been a pretty accurate way to “guess” wins.
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/13328023/Pythag.png)
According to their math, the Colts are playing like a team with 3.7 wins (8.4 on the season), and ranks 16th of all teams this year.
(EDIT: altered language to convey that 8.4 wins is not a prediction but a production level comparison)
BLOWOUTS
Blowouts are a key indicator of how good a team is, in that they should give them and not receive them. The Colts have done neither as all of their games have been within 1 score. I defined a blowout as a game with a point differential greater than 2 scores (16 points) and applied a +1 value for the winning team and -1 value to the losing team. Then, I ranked teams by their “net blowout games”.
.
| Rank | Team | Wins | Blowout loss | No Blowout | Blowout Win | Net Blowouts | Games | % Blowouts |
|---|---|---|---|---|---|---|---|---|
| Rank | Team | Wins | Blowout loss | No Blowout | Blowout Win | Net Blowouts | Games | % Blowouts |
| 1 | NE | 8 | 3 | 5 | 5 | 8 | 62.5% | |
| 2 | SF | 7 | 4 | 3 | 3 | 7 | 42.9% | |
| 3 | DAL | 4 | 4 | 3 | 3 | 7 | 42.9% | |
| 4 | MIN | 6 | 5 | 3 | 3 | 8 | 37.5% | |
| 5 | KC | 5 | 6 | 2 | 2 | 8 | 25.0% | |
| 6 | LA | 5 | 6 | 2 | 2 | 8 | 25.0% | |
| 7 | BAL | 5 | 6 | 1 | 1 | 7 | 14.3% | |
| 8 | GB | 7 | 7 | 1 | 1 | 8 | 12.5% | |
| 9 | SEA | 6 | 7 | 1 | 1 | 8 | 12.5% | |
| 10 | HOU | 5 | 7 | 1 | 1 | 8 | 12.5% | |
| 11 | TEN | 4 | 7 | 1 | 1 | 8 | 12.5% | |
| 12 | LAC | 3 | 7 | 1 | 1 | 8 | 12.5% | |
| 13 | NO | 7 | 1 | 6 | 1 | 0 | 8 | 0.0% |
| 14 | IND | 5 | 7 | 0 | 7 | 0.0% | ||
| 15 | CAR | 4 | 1 | 5 | 1 | 0 | 7 | 0.0% |
| 16 | JAX | 4 | 8 | 0 | 8 | 0.0% | ||
| 17 | PHI | 4 | 2 | 4 | 2 | 0 | 8 | 0.0% |
| 18 | CHI | 3 | 7 | 0 | 7 | 0.0% | ||
| 19 | DET | 3 | 7 | 0 | 7 | 0.0% | ||
| 20 | PIT | 3 | 1 | 5 | 1 | 0 | 7 | 0.0% |
| 21 | TB | 2 | 7 | 0 | 7 | 0.0% | ||
| 22 | DEN | 2 | 1 | 7 | -1 | 8 | -12.5% | |
| 23 | BUF | 5 | 1 | 6 | -1 | 7 | -14.3% | |
| 24 | CLE | 2 | 2 | 4 | 1 | -1 | 7 | -14.3% |
| 25 | NYG | 2 | 3 | 4 | 1 | -2 | 8 | -25.0% |
| 26 | ATL | 1 | 2 | 6 | -2 | 8 | -25.0% | |
| 27 | WAS | 1 | 2 | 6 | -2 | 8 | -25.0% | |
| 28 | CIN | 0 | 2 | 6 | -2 | 8 | -25.0% | |
| 29 | ARI | 3 | 3 | 5 | -3 | 8 | -37.5% | |
| 30 | OAK | 3 | 3 | 4 | -3 | 7 | -42.9% | |
| 31 | NYJ | 1 | 3 | 4 | -3 | 7 | -42.9% | |
| 32 | MIA | 0 | 4 | 3 | -4 | 7 | -57.1% |
Using that math the Colts are tied for 13th place. I admit it’s a bizarre way to rank teams but you can see that there is a pretty clear distinction between teams that blow teams out and those that get blown out and the teams in the middle usually have bad to average records with the Colts and the Saints being the exceptions.
DRIVE SUCCESS RATE
Drive Success Rate is a very good predictor of future point differentials. The Colts have a season long offensive DSR of 77.0% which ranks 7th best, but the defense has given up a rate of 75.2% which ranks 25th. The DSR differential of 1.8% is the 13th highest of any team.
DVOA
So far, none of these measures have adjusted for opponent. Football Outsiders is famous for ranking teams by their DVOA metric, which includes an opponent adjustment. After week 8, FO ranks the Colts 16th.
CONCLUSION
These empirical measures of the Colts rank them as somewhere between the 13th and 16th best team. With such strong agreement between the various measures it is hard to ignore the implication. The Colts record is probably better than the actual play they have put forth.
The good news is that those 5 wins are already in the bank and it doesn’t matter how they were achieved. With the state of the AFC South, another 5 wins is probably good enough to get to the playoffs.
The bad news is that going forward, if the level of play does not improve, you can probably expect a regression to the mean win-wise and 5 wins may not be do-able. Since moving to a 16 game schedule, there has been 55 teams that have started the season with a point differential between +4 and +10 after 7 games. 29 of them (53%) finished the season with 4 or fewer additional wins. Even if the Colts do get in, I’m not seeing a playoff win at this level of play.
Of course, all of these numbers don’t account for injury, team improvement, future strength of schedule etc. So, the data might be completely irrelevant and off base (like a lot of you think stats in general are). Who knows? The future is hard to predict. Last year, I thought the team was better then their 2-5 record and they were. This year I think they are worse than their 5-2 record. We’ll see.