PECOTA was originally developed by Nate Silver, who has since gone on to far greater fame (and likely fortune as well) as a political number cruncher, with his website FiveThirtyEight.com. You don’t really need to get into the details of what makes PECOTA tick. But according to Baseball Prospectus, there are three elements to the system:
- Major-league equivalencies, to allow us to use minor-league stats to project how a player will perform in the majors;
- Baseline forecasts, which use weighted averages and regression to the mean to produce an estimate of a player’s true talent level;
- A career-path adjustment, which incorporates information about how comparable players’ stats changed over time.
Put it all together, and you get projections for each player. Allocate playing time, and you can work out overall team statistics, which can also be used to generate an expected W-L record. So, here are the standings they expect for the NL West.
- Dodgers: 98-64
- Giants: 86-76
- Diamondbacks: 78-84
- Rockies: 76-86
- Padres: 70-92
Interestingly, that 78 win mark is exactly the same as PECOTA projected for Arizona last season, though obviously that was before the loss of A.J. Pollock for the entire season. I wonder if anyone has asked Mike Hazen what he thinks of the projection? Because last season, then-GM Dave Stewart was less than impressed by the prediction, saying "Think we only win 78 games?. That’s a joke." His skepticism is also shown in the pre-season video below, after Steve Berthiaume brought it up. PECOTA talk starts at around the 1:50 mark.
In terms of change from 2016, the projection expects the main source of improvement to be our pitching, which is expected to concede 740 runs this season. That’s a whopping hundred and fifty fewer than they did this season. However, this is somewhat negated by the offense taking a bit of a step back - the loss of Jean Segura and Welington Castillo no doubt play a significant part there - with the D-backs scoring 708 runs, which is also down on last year, albeit by only 44. Tomorrow, I’ll get into the details, and break down the projections for the D-backs a bit further.
But you might be wondering, how accurate were the numbers from last year? For fun, here are all 30 teams, in descending order of actual wins (the first number), along with the pre-season PECOTA projection (the second number), and the difference, plus or minus, between reality and the projection (+ = team did better than expected)
- CHC - 103, 92, +11
- TEX - 95, 80, +15
- WSN - 95, 87, +8
- CLE - 94, 92, +2
- BOS - 93, 88, +5
- LAD - 91, 94, -3
- BAL - 89, 72, +17
- TOR - 89, 86, +3
- SFG - 87, 87, 0
- NYM - 87, 91, -4
- DET - 86, 79, +7
- SEA - 86, 84, +2
- STL - 86, 82, +4
- HOU - 84, 87, -3
- NYY - 84, 85, -1
- KCR - 81, 76, +5
- MIA - 79, 76, +3
- PIT - 78, 79, -1
- CHW - 78, 82, +4
- COL - 75, 74, +1
- LAA - 74, 76, -2
- MIL - 73, 77, -4
- PHI - 71, 65, +6
- ARI - 69, 78, -9
- OAK - 69, 75, -6
- ATL - 68, 68, 0
- SDP - 68, 70. -2
- TBR - 68, 91, -23
- CIN - 68, 74, -6
- MIN - 59, 79, -20
Obviously, some major whiffs there - most obviously, over-estimating the Rays, and under-estimating the Orioles. But for 18 out of the 30 teams, the projections were within five games of the actual mark achieved. So, based on that, there’s about a 60% chance the D-backs in 2017 will be somewhere between 73 and 83 wins.
What do you think? Is 78 wins too optimistic? Too pessimistic? Or about right?