New views on in-season BABS

We are 10 weeks in and the BABS database was updated over the weekend. Your survey results have gotten me to rethink how we use BABS during the season, so I’ve started to reimagine different presentations of the data.

First, some clarifications from recent reader emails. The “2017p” line in the database represents our pre-season expectations for each player. That “p” is misleading, I suppose, but it does not represent balance-of-season projections.

In truth, once you are past your draft, you probably should not be using those “2017p” ratings at all. It’s better to scan each player’s multi-year record of actual performance and analyze the historical ratings to create their own picture. For instance, take Charlie Blackmon:

Year   Tm    AB     HR    SB   Avg   OBP   PT    Pw     Sp     Av
2014   COL   593    19    28   288   335    F    p      s      AV
2015   COL   614    17    43   287   347    F    p      S+     AV
2016   COL   578    29    17   324   381    F    PW            A+
2017   COL   256    14     5   336   380    F    PW     s      A+

2017p  COL                                  F    p      s      AV

His 2014-2017 scan tells a story of improving power and average, and signs of some latent speed. His 2017p line drew some conclusions from past data, but clearly confuses things right now.

Moving on, I have been playing with the asset group rankings based on your input. I was curious to see how the groups would be constructed if we lifted the playing time constraint. Here are the top 50 hitters according to their currently ranked asset groups. The asset groups are all separated and the players within each group are ranked by their current roto values.

The early listed players are in their own asset groups; the (P+,A+) group is the first one worth analyzing. These six players range in current earnings from $18 to $44, with the group’s mean value at around $33. Paul Goldschmidt’s value is inflated by his stolen bases, which BABS continues to neglect, but four of the six are earning right around the mean.

That leaves Matt Kemp as an interesting trade acquisition candidate. His low current earnings might lull owners into undervaluing his skill, but BABS sees him as having higher upside.

Liabilities are not shown here, but we still cannot ignore Harper’s (inj-) injury risk, and more important, Judge’s (EXP) experience risk. So maintain perspective on the sustainability of that duo’s current performance levels.

The (PW,AV) group has a similar earnings range and potential trade opportunities. Mike Moustakas (pictured) and Jonathan Schoop could be low-cost trade targets. The (P+,a) group features the overachieving Miguel Sano (note his batting average compared to the others in that group), who could be used as prime trade bait. Hernan Perez and Brandon Drury are two potential targets in the (p,AV) group.

The top 50 pitchers, according to BABS:

Low-inning closers always tend to put up better raw skills because stamina is not an issue. That makes Alex Wood’s appearance in the (E+,K+) asset group all the more remarkable. The seven closers who make up the rest of the group are all superb, no matter how many save opportunities their teams have given them and how many roto dollars they’ve earned so far.

The (ER,K+) group hold some interesting opportunities. Pass on the big three, and that leaves Robbie Ray and Lance McCullers as comparable talents. Trevor Cahill has pitched half as many innings, which makes his ratings a bit more suspect, but he’s managed some elite skills in the 41 innings he has pitched.

All of the other multi-player asset groups offer some potential trade targets and trade bait. The wider the roto earnings spread, the more interesting the opportunities. For instance, Carlos Martinez, Luis Severino and Drew Pomeranz are interesting targets given the better players in their respective asset groups. But the (ER) group features Ervin Santana, whose surface stats and $38 earnings far exceed the other comparable talents there. It’s tough to see his good fortune lasting much longer.

Let me know if these types of analyses are worthwhile. If so, we’ll start transitioning, though we’re probably looking at 2018 before we can implement all these changes in the database.