There are many challenges when ranking prospects. Two different people can see the exact same player, and depending on the preference of the individual people, they can come to completely different conclusions about the player’s upside.
As an example in the Pirates’ system, Ke’Bryan Hayes is a strong defender at third base who shows good contact ability and plate patience, while also adding some speed on the bases. He doesn’t hit for much power right now, but is young and has a projectable frame, and could add that power in the future. If you value defense at third, along with a good ability to hit, then you don’t need the power as much, and Hayes ranks higher in your view. If you think there should be more power from third base, and don’t value the defense as high, then you would be lower on Hayes.
Either way, Hayes is the exact same player. It just depends on how you value him. I should note that in our latest rankings, everyone had the exact same rankings for Hayes on his floor, ceiling, and likely upside, which is rare. Even top overall prospect Mitch Keller had some differences between all of the rankings. So Hayes might not have been the best example from our rankings to show a difference, but he is a good example to highlight how different preferences can alter rankings from person to person.
Those are some of the challenges with ranking individual players. But what about the challenges when combining the overall players? How do you compare a third base prospect like Hayes with a starting pitching prospect like Taylor Hearn, or a first base prospect like Will Craig? Or, as an even bigger challenge, how do you compare a third base prospect in High-A with projection versus a relatively safe pitcher like Steven Brault in the upper levels, who has a lower ceiling, but much less risk?
We’ve adjusted our approach to this over the years, trying to find the best approach to balancing floor, ceiling, and positional rankings. For several years, we’ve used “Likely Upside” as a big factor, which usually comes in below the ceiling, but above the floor. As an example, Mitch Keller’s average floor in our rankings was just below a 5.0, which means we think he’s going to at least be a #3-5 starter, or a closer candidate. His average ceiling was a 7.0, meaning we all thought he has the ceiling of an All-Star pitcher. But his likely upside comes in between those two, just below a 6.0, which puts him as a strong #3 innings eater.
That doesn’t sound appealing for your top overall prospect, especially when a team like the Pirates needs Keller to be a top of the rotation guy. But the “Likely Upside” rating is meant to be conservative, factoring in some risk to drop the ceiling.
There was another problem with the Likely Upside approach that we found over the years. Two players could still have the same Likely Upside, but have much different floors and ceilings. So we took things a step further this year and created a SUM rating (scientifically named after the function in Excel). We took the averages of the Likely Upside, Floor, and Ceiling rankings, then added them together. It’s not the most scientific approach, but it factors in the floor and ceiling once more.
The result is that a player with a floor of 2.0 (career minor leaguer), a ceiling of 5.0 (average player), and a likely upside of 3.0 (up and down player) would be ranked higher than a guy in Triple-A who is strictly an up-and-down player (3.0 Likely Upside, with a 3.0 floor and 3.0 ceiling). Both players have a 3.0 LU, but the projectable player (likely in the lower levels) has a higher ceiling, and his SUM rating of 10 would put him above the Triple-A depth player.
This approach has actually worked out very well, sorting out the raw rankings in a way that has led to a lot less discussion about specific rankings, as it puts the players in a better order. But there is also an added benefit: we can easily compare the system rankings over time by looking at the SUM rankings. So that’s what we did, comparing the 2017 Mid-Season rankings with the 2017 pre-season rankings, in order to see how the system has changed. Here are some quick comparisons.
The Overall System
We ranked 211 players in the mid-season rankings, compared to 207 in the pre-season rankings. Neither list included players who had yet to make the jump to the United States. When we took the total of the SUM rankings, we got the following:
The mid-season rankings did have a few more players, but not enough to make a difference in these results. The overall rankings show that the mid-season totals were higher. It’s not a significant difference, and I’m not sure you’d see such a massive change over a short period of time. But it’s good to see that the overall system isn’t declining.
The Top of the System
Naturally, we’d want to look at how the top of the system compared. The problem here is that we prefer tiered rankings, and comparing the top 10 or the top 50 would go against that. The pre-season rankings had 11 players in the top three tiers. The mid-season rankings had 13. So let’s break it down tier by tier.
This one took a huge hit, as Josh Bell and Tyler Glasnow both graduated. Mitch Keller and Austin Meadows are the remaining players from both tiers, and dropped in our mid-season rankings. This was due to the hamstring from Meadows, and some injury concerns from Keller, plus the fact that our mid-season rankings tend to be more conservative.
This isn’t a fair comparison, as the second tier at the start of the season was just Kevin Newman. He was joined by four other players at mid-season, although this was a case of Newman dropping a bit in value, rather than the other players increasing their value. Newman’s mid-season ranking would have clearly put him in tier 3 during the pre-season, which would have made that tier 2.
There were six players in each tier. However, as mentioned above, the third tier in the pre-season rankings was a bit stronger because Newman was stuck between that tier and the top guys. Three of the players from the pre-season list moved up to Tier 2 at mid-season. There were also some players who moved up to tier 3 from the pre-season rankings, such as Elias Diaz, Calvin Mitchell, Braeden Ogle, and Edgar Santana.
We’re getting to about 20 prospects here, with the pre-season list now including 19 players, and the mid-season list including 22 players. The top of the pre-season tier 4 would have ranked in tier 3 at mid-season. Again, we’re seeing the effects of Newman being in his own tier. However, this is where things start to normalize, as several of the mid-season tier 4 players would have also been in tier 4 during the pre-season, based on the comparison of rankings.
We’re now getting down to prospect number 30, with the mid-season list stopping at 31, and pre-season stopping at 30. There are two more players in the pre-season tier. The rankings continue to normalize here, as players in the mid-season rankings would have likely ended up in the bottom half of the pre-season tier. In looking at the lists, I’ve noticed that the pre-season tiers at this point are about half a step above the mid-season tiers, due to the Newman impact.
For some reason, tier 6 had seven players in each set of rankings. Things are starting to balance out here, as the tiers are getting to be about the same.
We typically stop tier 7 with players who have a SUM of 9.0. That amounts to a few common rankings, like 3/2/4, which means that a player’s likely upside would be an up and down player, his floor is a career minor leaguer, and ceiling is a bench player. Jin-De Jhang was an example of this ranking in our latest list. Jacob Stallings is about the same, although he’s in the 3/3/3 category, where he’s an up-and-down player already, and won’t likely exceed that. Usually the 9.0 guys are the ones who have a shot to make the majors. Guys below that could still make it, but right now the risk is too high to include them, or consider them a strong bet to be more than a career minor leaguer.
The Top 50+
The rankings are about the same, with 67 players in the pre-season rankings, and 70 in the mid-season group. That might give a bit of an edge to the pre-season group, but a disclaimer comes into play.
One thing to consider is that the Pirates graduated some top prospects, and the players who graduate typically have some of the highest rankings. The big four were Bell, Glasnow, Trevor Williams, and Jose Osuna. That removed 61.20 points from the system, based on the pre-season rankings. There were other players who left the system, such as Alen Hanson, Lisalverto Bonilla, Frank Duncan, Tyler Webb, and Brady Dragmire, to name a few, although we’re always just one waiver claim or minor trade away from getting Dragmire back. Again.
The sum of the players who left from the top 7 tiers was 57.20. That means the pre-season list had 118.40 points that the mid-season list had to make up. The Pirates added 13 new players to the top 7 tiers, although that comes with the disclaimer that two of those players (Lolo Sanchez and Sergio Cubilete) were previously in the system, but didn’t get graded as players in the DSL. Those 13 players amounted to 138.32 points.
So the Pirates lost a lot of talent from the system, mostly through graduates, but also through the typical amount of players who are traded away or who get waived. They more than replaced that talent with new draft picks and players coming up from the DSL. The pre-season losses were more top-heavy, while the mid-season additions are more evenly spread out through the system.
The Returning Players
Once we remove players who have graduated and left, and players who were added mid-season, we can compare the top seven tiers on the basis of players in the system for both lists.
The edge goes to the pre-season group here. The groups include different players, but as an overall picture, the top ranked players in the system saw a slight decline from pre-season to mid-season. This was balanced out by the new additions to the system outweighing the players who graduated or left.
The Overall Trend
The overall system is slightly up, while the top seven tiers are slightly down. Essentially, the system is in the same place as the pre-season, with a slight edge to the pre-season group. This is actually kind of impressive, considering the Pirates graduated two of their top prospects, along with two more talented prospects, and lost several others from the pre-season list. They were able to replace the overall value, although the system isn’t as top-heavy as it was pre-season.
That’s the big thing I noticed when browsing the rankings. The system lost value at the top, but added a lot more options in the middle and bottom. The good thing about this is that most of the guys who increased in value, or who were added to the system rankings, are young players who are highly projectable. This comes from a prep-heavy draft class, plus a good group of international prospects making their way up in the low levels.
The system almost reminds me a bit of the 2011-12 time frame, where the Pirates had a lot of highly projectable players in the lowest levels. Those players went on to make the Pirates a top farm system, and many of them are playing big roles in Pittsburgh right now. If the Pirates can get a similar result from this current group, they could see another top system in a few years, with another top-heavy group of prospects.