I assume this quote is from the Athletic article? It seems simplistic to me. As someone who creates forecasting models for a living (and therefore understands the difference between a forecast and a projection), I don’t think one can compute a “margin of error” for such a projection.
If the author is basing that statement on something, it’s probably past performance of PECOTA in projecting the number of wins for a team. I am not familiar with the details of how PECOTA works, but my limited understanding is that it projects individual performance and sum’s the results to estimate how the team is expected to do. I can think of three major things that can affect these projections:
- Players performing better or worse than projected. This happens all the time, of course, sometimes due to injury.
- Roster composition changes, due to injuries, trades, promotions from the minors, players being DFAed, etc. Also happens a lot.
- Players being used differently than projected.
The manager may have only limited control of #1 and #2 but is the major influence on #3. It may be hard to separate the effects of #1 and #3, and so the manager’s impact is hard to quantify.
I don’t know what was projected by PECOTA for Giants players for 2023, but I suspect that some players performed much worse than projected, for example Brandon Crawford and Joey Bart. That underperformance seems likely due to #1 and is probably not Kepler’s fault. But I don’t know enough to excuse him regarding good players like Wade and Flores. Could they have been used more optimally?
The other issue I have here is treatment of the as a whole. The Giants were 13 games over .500 when the season was 60% over; now they’re four games under. If the Giants are truly a .500 team, they should have been much better the last two months. If the author wants to credit Kapler for bringing a .500 team in at .500, does Kapler deserve blame for a .500 team losing two thirds of their games for a third of the season?