I'm ahead of the game when it comes to prepping for the draft (consider it procrastination from actually writing my dissertation) and I noticed some interesting things when looking at how different projection systems value players. I'm going to be a bit vague here and there, so as to not give too much of my board away, but there are certainly some interesting tidbits here. To be clear, I'm working with the exact same set of constraints regarding hitter/pitcher splits, how to handle bench players, and other similar decisions, so differences there aren't driving these values.
Steamer and ESPN say top pitching is a worthwhile splurge. Fans and Baseball Prospectus say pitching is deep (for different reasons)
According to projections from Steamer and calculations available to ESPN Insiders, the top 10 pitchers should be rather highly paid - an average of $26 and $25, respectively. This matches up with what we spent on the top 10 pitchers last year - $26/top ten arm (including David Price, who was kept at $25). Fangraphs Fans and Baseball Prospectus' Player Forecast Manager, on the other hand, had the top 10 arms much more lightly valued, both coming in at $21 on average. What is interesting is that the Fans have this for two reasons. First, they project many young, rising pitchers to be excellent and surprise players to maintain their excellence. For example, both Sonny Gray and Hisashi Iwakuma both sneak into the back of the top 10. Second, by having very good projections for players who are expected to be back of the fantasy rotation arms, that substantially raises the replacement level, and therefore knocking down value off the top. Third, the fans are projecting several truly dominant performances by closers. Based on the projected statistics, Fans are valuing the top 5 closers at $25 on average; put another way, a top closer is worth more than a top starter. This is really being driven by some incredibly low ERAs and WHIPs, coupled with some spectacularly high save totals. Because the money in the pitcher pool is a fixed amount, by paying closers for their incredible projected performances, you're taking money away from starters.
Baseball Prospectus, on the other hand, ends up with the conservative valuation of top tier pitchers through more traditional and expected routes - conservative projections. All told, the BP projections are more conservative across the board, presumably due to the incorporation of regression to the mean into PECOTA.
ESPN is ready to bury high power, low average players
Two of the biggest differences for the valuation of individual players come from Adam Dunn and Nick Swisher. Both are fairly highly valued by Steamer projections, but barely rosterable per ESPN's projections.
Taking a closer look at Nick Swisher's projection, things get even more interesting...
These projections are almost as close as you can possibly imagine, with four of the categories being exactly the same. So just how do you end up with a $16 gap between the two identical projections?
Baseline.
ESPN's projections are very, very optimistic in terms of production, while Steamer's are much more conservative. For example, Swisher's .255 average is almost 20 points below the average for a starter in the league per ESPN. That exact same .255 average is only 10 points below average per Steamer. Ditto for home runs - Steamer has the starters hitting 18.3 home runs on average, while ESPN has them hitting 19.7. The Swisher example is a particularly striking one, but this is a great illustration about how these projections are apple and oranges, at least with regard to individual player lines.
The fans LOVE their bright new shiny baubles
Consistently across the board, fans rate unestablished, but up and coming players much more favorably than any of the other three projection systems.
Check out these high upside outfielders and where they rank at the position:
Avisail Garcia, for example, actually ranks higher than Jacoby Ellsbury and Carlos Beltran, Moss and Calhoun are ahead of Alfonso Soriano and Austin Jackson, and Oswaldo Arcia is ranked ahead of Ben Zobrist, Desmond Jennings, and Brett Gardner. It is important to recognize that the fans don't seem to be taking risk or downside into account properly, both in terms of performance and playing time. Perhaps when someone thinks "Kole Calhoun could hit 25 HR and steal 15 SB next year", that's what they put in as their projection. Another important aspect to this is that people are putting in projections, not rankings. I think if you asked anyone who put even the most optimistic Avisail Garcia projection, if they'd rather have Ellsbury or Garcia, they'd look at you like you had two heads. Nevertheless, that is how the projections work out. Either way, it is incredible to see players who will probably be a few bucks ranked ahead of $25+ players like Ellsbury.
Steamer and ESPN say top pitching is a worthwhile splurge. Fans and Baseball Prospectus say pitching is deep (for different reasons)
According to projections from Steamer and calculations available to ESPN Insiders, the top 10 pitchers should be rather highly paid - an average of $26 and $25, respectively. This matches up with what we spent on the top 10 pitchers last year - $26/top ten arm (including David Price, who was kept at $25). Fangraphs Fans and Baseball Prospectus' Player Forecast Manager, on the other hand, had the top 10 arms much more lightly valued, both coming in at $21 on average. What is interesting is that the Fans have this for two reasons. First, they project many young, rising pitchers to be excellent and surprise players to maintain their excellence. For example, both Sonny Gray and Hisashi Iwakuma both sneak into the back of the top 10. Second, by having very good projections for players who are expected to be back of the fantasy rotation arms, that substantially raises the replacement level, and therefore knocking down value off the top. Third, the fans are projecting several truly dominant performances by closers. Based on the projected statistics, Fans are valuing the top 5 closers at $25 on average; put another way, a top closer is worth more than a top starter. This is really being driven by some incredibly low ERAs and WHIPs, coupled with some spectacularly high save totals. Because the money in the pitcher pool is a fixed amount, by paying closers for their incredible projected performances, you're taking money away from starters.
Baseball Prospectus, on the other hand, ends up with the conservative valuation of top tier pitchers through more traditional and expected routes - conservative projections. All told, the BP projections are more conservative across the board, presumably due to the incorporation of regression to the mean into PECOTA.
ESPN is ready to bury high power, low average players
Two of the biggest differences for the valuation of individual players come from Adam Dunn and Nick Swisher. Both are fairly highly valued by Steamer projections, but barely rosterable per ESPN's projections.
| Difference | High Proj. | Low. Proj. | |
| Adam Dunn | 19.46 | Steamer | ESPN |
| Nick Swisher | 15.7 | Steamer | ESPN |
Taking a closer look at Nick Swisher's projection, things get even more interesting...
| HR | R | RBI | SB | AVG | |
| Steamer | 23 | 80 | 85 | 2 | 0.255 |
| ESPN | 23 | 80 | 79 | 2 | 0.255 |
These projections are almost as close as you can possibly imagine, with four of the categories being exactly the same. So just how do you end up with a $16 gap between the two identical projections?
Baseline.
ESPN's projections are very, very optimistic in terms of production, while Steamer's are much more conservative. For example, Swisher's .255 average is almost 20 points below the average for a starter in the league per ESPN. That exact same .255 average is only 10 points below average per Steamer. Ditto for home runs - Steamer has the starters hitting 18.3 home runs on average, while ESPN has them hitting 19.7. The Swisher example is a particularly striking one, but this is a great illustration about how these projections are apple and oranges, at least with regard to individual player lines.
The fans LOVE their bright new shiny baubles
Consistently across the board, fans rate unestablished, but up and coming players much more favorably than any of the other three projection systems.
Check out these high upside outfielders and where they rank at the position:
| Difference | High Proj. | Low. Proj. | Rank | |
| Avisail Garcia | 17.19 | Fans | Steamer | 9 |
| Brandon Moss | 18.7 | Fans | BP | 12 |
| Kole Calhoun | 14.55 | Fans | ESPN | 14 |
| Oswaldo Arcia | 12.51 | Fans | ESPN | 19 |
Avisail Garcia, for example, actually ranks higher than Jacoby Ellsbury and Carlos Beltran, Moss and Calhoun are ahead of Alfonso Soriano and Austin Jackson, and Oswaldo Arcia is ranked ahead of Ben Zobrist, Desmond Jennings, and Brett Gardner. It is important to recognize that the fans don't seem to be taking risk or downside into account properly, both in terms of performance and playing time. Perhaps when someone thinks "Kole Calhoun could hit 25 HR and steal 15 SB next year", that's what they put in as their projection. Another important aspect to this is that people are putting in projections, not rankings. I think if you asked anyone who put even the most optimistic Avisail Garcia projection, if they'd rather have Ellsbury or Garcia, they'd look at you like you had two heads. Nevertheless, that is how the projections work out. Either way, it is incredible to see players who will probably be a few bucks ranked ahead of $25+ players like Ellsbury.
