With PECOTA projections out, let us look at Blue Jays hitters ceilings and floors based on current projections.
It is the time of year where everyone is predicting what the upcoming MLB season will give us. Some love to speculate on anecdotal evidence, some use statistical evidence, some rely on computer-modeled projections. There are many forms of computer models out there, there are 4 or 5 on Fangraphs alone.
My favorite projection system to use is PECOTA from Baseball Prospectus. PECOTA stands for Player Empirical Comparison and Optimization Test Algorithm. It is one of the original computer-modeled projections out there, created by Nate Silver, back in 2003.
For most of PECOTA’s history, the analysis was based on 50th percentile outcomes; i.e. the most likely outcomes. It runs thousands of simulations based on past statistical inputs and the median outcome would be the 50th percentile outcome.
New this year, PECOTA has added a new feature that allows us to look at 99th percentile (the ceiling) outcomes and 1st percentile (the floor) outcomes. We will also look at most likely outcomes at the end to put the numbers into context.
Today we will be looking at the Blue Jays hitters and what their range of outcomes could be for this year. Keep in mind, the range of outcomes goes from the 1st percentile to the 99th percentile, meaning these numbers should capture all possible outcomes, but there is a slim possibility that they are wrong.
1st Percentile Outcomes
Here is the 1st Percentile Table.
This is not something we really want to see, is it? This is the absolute floor for every hitter on the Blue Jays. I have stripped out some players that will likely not see more than 50 plate appearances. Imagine a world where every player on the Jays had their worst possible season in 2020, this would be it. Not a player with a 0.300 OBP and one player batting above .200. That is pretty grim. These numbers all assume the player stays healthy for most of the season.
A couple of positive things of note from this chart.
Cavan Biggio‘s floor for 2020 is still a 1.5 WARP (this is Baseball Prospectus’s version of WAR), with 4 FRAA (Fielding Runs above average, a similar stat to DRS or UZR), and a 87 DRC+(Deserved Runs Created, similar to wRC+ but with deserved outcomes on batted balls). Biggio, regardless of how poor we consider his season, is still nearly league average at providing value to the Blue Jays.
Reese McGuire‘s floor for 2020 is a positive WARP season with excellent defense. I guess we could have expected this, as we know, defense is not something that usually slumps and is fairly consistent. So even with poor offensive numbers, McGuire will still provide value.
Also no matter what your feelings are towards Randal Grichuk, he should still provide 22 home runs with his worst possible outcome.
99th Percentile Outcomes
Here is the 99th percentile table.
Doesn’t this table paint a rosy picture of the Blue Jays? No player under 100 DRC+ (all above league average offensively), most players above .400 OBA, 11 players having 2 or more WARP, and 8 guys with 30 or more home runs. Now, I am not a total dummy, I understand that this is a completely far fetched idea that all players have their best possible outcome all in the same year, but it is a fun idea to dream isn’t it?
Some things I noticed,
Cavan Biggio has the best WARP outcome in both the 1st and 99th percentile. He is an above-average defender with a high floor, but he also has an incredible ceiling too. His ceiling is a 5.8 WARP player, with 37 HR, 32 doubles, a 145 DRC+ all while slashing .335/.466/.659. That is incredible. This is merely the projection of his best possible outcome in 2020. He is still young and could improve as well. Do I think he will have this incredible season? Probably not, but it is in his range of outcomes.
Both Vlad Guerrero Jr. and Bo Bichette become 4+ win players in the 99th percentile. Both would derive most of their value from their bats. Although Bichette would get some value from his running abilities and fielding abilities. If the Blue Jays intend on being competitive in the next season or two, I believe Vlad and Bo will need to be 4 win players. Keep in mind, being a 4+ win player may not be their 99th percentile outcome next year.
PECOTA really seems to like Reese McGuire over Danny Jansen. On both the ceiling and floor of their projections, Reese provides more WARP. While I don’t agree with this, their model seems to have McGuire playing more often than Jansen. This could be the reason McGuire is providing more value in their projections.
If we tallied up all of the WARP of the hitters we would end up with 40.3 WARP. In this fictitious world, the Blue Jays would have a nearly 90 win team without even adding the pitching. In theory, a replacement-level team would win 48 games, add the 40+ wins from before, the Jays would be around 88 wins.
50th Percentile Outcomes
Here is the 50th percentile table.
Enough with the exaggerated scenarios. In the real world, some players outperform their average projections and some perform below, which is why looking at the 50th percentile outcomes, as a whole, could give you an idea of what this team could really do.
The projections really seem to like Reese McGuire and Cavan Biggio still. It also predicts that the Blue Jays have 7 players above 1 Win. It also only looks like they have 5 players hitting at a rate above a league-average DRC+. The team average of 95 DRC+ means they may behind the league average in offense.
The total WARP provided by the position players should be around 15.6. I expect, and this is just anecdotal, the hitters to surpass this total. PECOTA does not do a great job at predicting breakouts or improvements and since most of the important players on the Blue Jays are still very young, they have yet to break out.
PECOTA relies on the previous three years of performance to predict the future. It cannot take into account any changes a player has made because that stuff is unknown. For example, If Travis Shaw has made some changes with the goal of getting back his 2018, line drive, swing, PECOTA will not know about it. Last year the Jays hitters combined for 16.8 WARP, I believe they can get to this level if not higher.
Range of Outcomes
Since WARP is a cumulative statistic, the more one plays the more chance one has to affect the outcome of the stat. The players with the largest ranges in outcomes tend to be the ones who play the most. Also of note, the players with the least amount of service time are near the top in range of outcomes as well. Guerrero, Bichette, and Biggio all of have a range of outcomes larger than 4.3 WARP or more. The more of a track record a player compiles, the smaller this range should be.
Lastly, I would like to say, these are just projections from a computer model, they are not something set in stone, they are merely just predicting the most likely outcome based on their data. If a player outperforms the model, that is awesome, and will just provide more data on the player for the future. I love using PECOTA as a guideline as to how well the player is playing withing his talent level.
I will put out pitching ceilings and floors based on PECOTA projections very shortly.