Projecting the Defensive Future of Minnesota's Top Prospects

[Mark G]

1 hour ago, Major League Ready said:

The measure is Wins Above Replacement.  Given 1B produces more offensive production than 2B the replacement player at 1B needs to produce more than someone at 2B to have the same impact on WAR.  Once again, you think someone else is an "idiot" (your words) because you don't understand this analytical construct.  I would bet the people who developed that stat have formal educations in quantitative measures and other credentials that actually got them that job.  Yet, if you don't understand, you assume the problem is that the people who actually have the credentials to have this job are idiots.  

All of this is true to an extent.  But just a personal anecdote here: I spent 10 years of my life working at a couple of Universities/colleges where some of the professors had IQ's higher than Jake Caves batting average.  But get them out of the classroom and off the blackboard and they could barely find their way through a grocery store.  Brilliance does not always equal common sense, and I wonder how many of the analytical geniuses that create the value we now put on certain player skills really understand the need for balancing a roster based on all different talents.   For every Sano, or any other player whose primary worth is power, there needs to be an Arraez to be on base to make that power productive.  A Simmons, for example, might make Donaldson a better 3rd baseman based on his range; same as Buxton has carried our left fielders for some time now with his range.  Is that factored into the WAR equation?  Is it factored into the monetary worth of each skill set?  If Sano's 30 HR's only equal 54 RBI's because no one is on base, is that worth more than Arraez scoring 74 runs by being on base as often as he is? (made up numbers, but you get the point)  Just because the slugging percentage is higher, is it worth more?  Or is it not Sano's fault no one is on base when he hits, therefore the worth is fixed?  Not an argument, just wondering if everything possible is taken into account when the analytics are compounded?   I have never totally understood the formula.

I do have one other question:  when it figures Wins Above Replacement, what constitutes a replacement?  Another player in your organization you would put there, or replacing a player with a player from another team through trade or FA?  Is that a hard and fast equation, or does it depend on the circumstances?  Or is it just the major league average for whatever position you are looking at?   That is one area I still don't understand.  Thanks.  

[ToddlerHarmon]

4 hours ago, Mark G said:

Not an argument, just wondering if everything possible is taken into account when the analytics are compounded?   I have never totally understood the formula.

To answer the specific question of whether on-base percentage is weighed correctly versus slugging when coming up with WAR, the answer is: there was an attempt.

WAR uses Runs Created (RC) as its basis for offensive production, which gives more weight to getting on base than OPS does. It also gives more weight to singles versus walks (because walks don't drive in runners from 2nd), credits steals, etc. So, WAR tries to fairly account for the different offensive skill sets.

Now, whether it does so accurately is another issue. The WAR calculation assumes a "normal" offensive environment, so a lineup with extreme power and poor on base skills might struggle more than WAR accounts for, and so might a lineup with good on-base skills and little power.

The other thing WAR does is assume that runs make wins. Which equates consistent production with streaky production, which is probably not accurate. A team that averages 4.5 runs a game by having a week of 1, 2, 2, 3, 8, and 10 runs probably went 2-4, but a team that scores 2, 3, 4, 5, 5, and 7 runs probably went 3-3 or 4-2.

So, IMO, WAR undervalues on-base skills, because something that happens 35% of the time (getting on base) is more consistent than something that happens 10% of the time (extra-base-hits).

[wabene]

On 10/21/2021 at 11:12 AM, Tim said:

a banjo hitter ? what the hell is that hahaha .. are guys that put the ball in play with a .313 / .374 all of a sudden players you don't want in the lineup? 

Aaron Gleeman just dropped his chopsticks

[TheLeviathan]

On 10/21/2021 at 1:15 PM, stringer bell said:

According to BBRef, Arraez profiles pretty close to average as a defender. He has positive numbers in both Total Zone Fielding Runs Above Average and BIS Total Runs Saved Above Average at second base in both 2020 and 2021. He was one run below average in Total Zone Fielding at third in 2021 and zero runs in left field. The BIS numbers are better--three runs saved in left and five at third. Overall Arraez' defensive numbers look decent except for his rookie year. 

I'll edit to add that Arraez is far from a complete player. He lacks power and speed and has been injured a fair amount, but in the batter's box, he sees a lot of pitches, makes contact at a high rate and gets on base more than any other Twin. 

Well said.  His skill set is rare these days.  Keeping him around will help this team but it's past time to give him a defensive home to get comfy in.

[Mark G]

6 hours ago, ToddlerHarmon said:

To answer the specific question of whether on-base percentage is weighed correctly versus slugging when coming up with WAR, the answer is: there was an attempt.

WAR uses Runs Created (RC) as its basis for offensive production, which gives more weight to getting on base than OPS does. It also gives more weight to singles versus walks (because walks don't drive in runners from 2nd), credits steals, etc. So, WAR tries to fairly account for the different offensive skill sets.

Now, whether it does so accurately is another issue. The WAR calculation assumes a "normal" offensive environment, so a lineup with extreme power and poor on base skills might struggle more than WAR accounts for, and so might a lineup with good on-base skills and little power.

The other thing WAR does is assume that runs make wins. Which equates consistent production with streaky production, which is probably not accurate. A team that averages 4.5 runs a game by having a week of 1, 2, 2, 3, 8, and 10 runs probably went 2-4, but a team that scores 2, 3, 4, 5, 5, and 7 runs probably went 3-3 or 4-2.

So, IMO, WAR undervalues on-base skills, because something that happens 35% of the time (getting on base) is more consistent than something that happens 10% of the time (extra-base-hits).

Spot on about the runs make wins point.  Our team this year finished 7th in the league in runs, but 13th in the league in wins.  That is because we scored 3 runs or less 45% of the games and many more runs than that in other games, win or lose.  

But what constitutes a replacement in the WAR?  Actually replacing the player, or simply taking the overall average from all players at that position?  And how would they figure the average, if WAR is what it is?  Thanks for your insight into all of this.  I still fall back on the old runs produced stat:  runs scored plus RBI's, minus home runs.  To me that is a great measure, and that is where Polanco shined.  162 runs produced through that formula in 152 games played.  Great stat.  Yet I only found a 4.8 WAR score for him.  That is why I really don't know how seriously to take WAR.

[ToddlerHarmon]

17 hours ago, Mark G said:

But what constitutes a replacement in the WAR? 

I have to admit ignorance of this part of the calculation. The idea of a replacement player is below average: your standard AAAA player - someone that can be obtained from waivers or a minor-league contract for the cost of nothing more than a roster spot. But that doesn't give me a number to hold on to. I've heard that a team of replacement players (0 WAR) would be expected to win about 48 games (again, with Runs -> Wins), but I don't know how that baseline gets set, either.

Polanco being credited with 4.8 WAR isn't at all a slight. It means a lineup of 9 2021-version-Polancos, saddled with the Saints rotation, bullpen, and bench, (think 5 ERA pitching and a bunch of .200-hitting benchwarmers) would grab 91 wins and a Wild Card berth, if not a division championship.

The usual benchmarks for hitters are 8 WAR is an MVP season, 5 WAR is an All Star season, 2 WAR is an average everyday hitter (for their position).

I'm sure you've heard it before, but the runs-produced stats are obviously dependent on other players' performance, and end up with clear outliers.

A famous example: 1985 Mattingly, a fine hitter, looks like Gehrig with 253 R+RBI, when in truth it was Rickey Henderson's INSANE production in front of him that made it all possible, but he "only" had 218 R+RBI, (which also doesn't credit the fact that Henderson played a much harder defensive position). WAR, which gives Mattingly all-star kudos at 6.5, and Henderson ALL-TIME recognition at 9.9 WAR, paints a more accurate picture.

[jorgenswest]

Does the changing environment of fewer balls in play, swings that lead to more fly balls and infields that shift with every batter also call for a change in the models we use to put a value on defense?

Is it possible that elite shortstop play doesn’t contribute as many wins as it did 10 or 20 years ago? Perhaps the Twins value defense at SS too highly in this new environment. Polanco might be the answer. 

 

[USAFChief]

If anyone really wants to know how the 2 primary WARS are calculated:

bWAR: https://www.baseball-reference.com/about/war_explained.shtml

 

fWAR: https://library.fangraphs.com/misc/war/

 

They are calculated differently. For example,  fWAR uses what a pitcher "should have done (FIP)," bWAR what actually happened (RA9). There are also differences in offensive measures used. The two systems use completely different measures of defense as well. 

 

Both, IMO, are well intentioned...and virtually worthless.

 

[mikelink45]

On 10/20/2021 at 7:54 AM, roger said:

Polonco at second, Lewis at short and Miranda at third.  Keep Buxton and Kepler in center and right.  Add Larnach and Martin in left field with Martin moving around to center and right as needed.  Arraez gets a lot of at bats as DH, while also backing up Polo and Miranda at second and third.  Kirilloff is the first baseman with Sano as DH and AK's backup.  

Would be heck of an every day lineup without the need to go out and sign any expensive free agents.  But it all starts with Lewis being able to handle short.  As for who plays short when Lewis needs a day off, could be Polo, Martin or another backup utility guy...maybe Gordon?  

I am in except for Kepler