As I posted on the SONAR dedicated page of the site, I’ve completed the offensive sheets for all 30 MLB teams, and you can find the links organized by division. If there are any errors with the links, please post in the comments section on the SONAR page. Now that I’ve finished all 30 sheets, I can start to look at the big picture, how the teams compare, and then begin to break it down position by position. I’ll also do a similar setup for all 30 teams with regard to the pitching side of things, before doing a cumulative rank of the 30 teams. After that, I’ll produce my top 15 prospects for each organization, which will be a combination of the SONAR scores as well as my own personal opinion based on the info I’ve gathered and my study of the prospects involved. So check below for more…
I’ll share three charts here, just basic stuff, and then kind of break down what I see.
The Phillies actually place last in terms of players with a positive score. A positive score is subjective, of course, because guys who scored 0.50 are counted in this reading. Seattle actually comes in tops, even after they pulled off a few trades of note. The Phillies lost two guys (Taylor, D’Arnaud) and gained one (Gillies), so they’d have had one more prior to the Halladay/Lee deals, but would have still ranked behind Tampa Bay for last spot. More surprisingly, the Mets came in 2nd, a team with a system that is very heavily criticized by the masses. Boston notably comes in 8th, and the Yankees in at 16th. There isn’t a whole lot of info to be gained from this, its just interesting to start off with.
This metric should be a bit more useful. This is simply a tally of all of the positive value players in each organization. The Padres (another surprise) lead the race with a score of 658.77, followed by the Yankees, Mets, Red Sox and Mariners. The Phillies come in 12th with a score of 442.08, just 2 points behind the Braves. Interestingly, all 5 NL East teams rank in the top 12, with the Mets 3rd (550.9), the Nationals 8th (478.19), the Marlins 10th (447.92), the Braves 11th (444.58) and the Phillies 12th. The Royals, who have been picking near the top of the draft for what seems like an eternity, come in dead last with a score of 218.76, some 30 points behind the Rockies. The average was just over 400, at 405.20, with 17 of the 30 teams scoring above average.
And finally, we have the system rankings based on the strength of the organization’s top 10 prospects according to SONAR. The Padres again lead the way, with a score of 481.23, more than 80 points ahead of second place Boston. The Nationals, believe it or not, come in 3rd with a total score of 390.69, followed by the Braves and Yankees. The Phillies actually rank 6th out of 30 teams here, with a score of 372.78. The average total score for the Top 10’s was 286.68, with 13 teams scoring above average. The Rockies actually have the lowest score, 158.75, with the Royals slightly ahead of them at 164.26. Based on all of their prospects, and then the top 10, SONAR doesn’t see a lot to love in those two systems. The Phillies are short in numbers, but make up for it in quality at the top. We’ll break down the other teams individually at a later date, this is simply a snapshot.
Phuture Phillies 
So the gist of it is this: If we make any more trades of prospects, our system will drop out of the top half very quickly as we have no depth?
Sounds about right.
LikeLike
The Astros and Cardinals rate highly here despite nearly everyone being in agreement that they have bottom five minor league systems? I’m sorry, but I really don’t see SONAR as anything but a junk stat.
LikeLike
Well I guess if that’s what the consensus says, then it must be true.
LikeLike
SONAR scores really aren’t at all useful unless we know what you’re taking into account that (seemingly) every other person on the planet ranking prospects isn’t.
LikeLike
Its been outlined in the intro piece which I linked above.
Its component based, and it focuses on performance. I’m not going to give you the formula, that wouldn’t make sense.
LikeLike
i just can’t trust something that tells me the opposite of what i believe to be true if i don’t at least know what it is being based on. sorry, this is the one thing that you do on this website that i don’t like.
LikeLike
What I find surprising and encouraging about this as Phillies’ fan is that it seems despite some pretty significant trades over the past 2 years (Blanton, Lee, Hallady), missing early round picks last year, and getting minimum production from early 2008 picks (Hewitt, Collier), the Phillies still have a pretty strong farm system (I give greater value to the top ten totals then the other cumulative stats). That’s pretty exciting.
I’ll be curious as you do team by team analysis how 1 great player like Heyward for Atlanta skews their ranking in this cumulative approach.
Note: I was posting as pete but since another Pete joined us, welcome other Pete, I changed to PRD for camaraderie’s sake.
LikeLike
Unlike Kevin and jpd, I really don’t care what the formula is. But you’ve posted 5 scores of past prospects, you missed on 4, and the one you did get, everybody else got as well. I need reason to believe that there’s actually some reason to think these numbers are of any value whatsoever. Until then, it just looks like a bunch of numbers thrown together.
LikeLike
Some may think the Phillies are ranked too low and the Astros too high, but the charts above are for hitting only. The Phillies do not have a bunch of huge bats in the minors. Brown, Gillies, Santana are Valle about it. The Phillies’ system strength is quality arms and projectable athletes. Not performance. Taken in that context, the charts do not seem off.
The SONAR scores for hitting, seem to be pretty reasonable to me. The formula focuses on some combination of .OBP, SLG, BB%, k% and age to level. With age to level weighted heavily.
LikeLike
I find it interesting. The Phillies offensively scored a little better than I thought they would given the trades plus the depth in the system is really in pitching. I think last year’s draft could ultimately surprise us if a few of the athletes like Hudson break through this year. Unlike many, I’m actually expecting a good year from Hewitt who should play more relaxed in the OF.
LikeLike
PRD: from what little I know about sonar, many of the prospects from last year’s draft wouldn’t be rated very highly. So the effect of the Phillies missing the first round last year wouldn’t catch up to them right away in terms of relative position.
That’s how I understand it.
LikeLike
Technically, not having a 1st round pick does effect their relative position. That is one less positive score for the system (assuming the pick performed at a positive level).
LikeLike
SONAR is crappa
LikeLike
I see the SONAR as a evaluation more then a projection tool and as such it is very good. For example Dom Brown if you do sonar for all his seasons played then you can see how he improved and as such you can evaluate him. His rookie season in the GCL he probably ended up in the negative Sonar score but improved every year since. Regardles so James should be commended for putting in a lot of effort in this. Comments like SidBream’s are really uncalled for.
LikeLike
These charts give the opposite conclusion of what we had seemingly ‘decided’ as a consensus view of the Phillies farm. At least I thought we had concluded that after the trades the Phillies 1) still had a great deal of depth, especially at the lower levels of the farm and 2) were sorely lacking in top talent, especially at the top of the system. These charts seem to say the opposite — that we have a very strong top 10 but are lacking depth. What am I missing, or is it just that pitchers aren’t included and that their inclusion will pivot the analysis?
LikeLike
This is obviously only one half of the equation, the position players side. The pitching half will obviously be 50% of the weighting of overall farm system strength
LikeLike
It seems to that the scores are a snapshot of the player right now. It really doesn’t take projections into account, just how they have done so far. Once pitching is calculated in, it will be interesting to see how the ranking end up. I believe it won’t be far off from other organizational rankings.
LikeLike
I refuse to accept that SONAR has any validity whatsoever until it aligns perfectly with the consensus.
Seriously, repeat after me, people: “data point.” As in, this is (like scouting reports, various statistical metrics, ARL, etc.) a data point that can prove useful if used in conjunction with other data points. Use some common sense — there’s nothing out there that tells you everything you need to know about a prospect, which is why you need to synthesize different pieces of information.
At first glance, SONAR’s biggest strength seems to be in identifying young guys in the lower minors whose components stand out, especially guys with ARL on their side. A weakness I see is that premium bats who don’t have ARL on their side don’t seem to score very highly. Understanding things like this can give us an idea of how to best use SONAR. It may take some time to grasp some of the intricacies, but I think it’s best to give it some time rather than dismiss it out of hand because it doesn’t fit within someone’s preconceived notions.
Just my two cents, anyway.
LikeLike
Best I can tell, here are the bigger issues with the statistic.
* We don’t know the exact formula of the statistic. That makes it terribly hard to review, and we’re just taking James’ word on a lot of it.
* As far as I can tell, it does not measure fielding statistics.
* It only measures one year of statistics at a time. 500-600 plate appearances is really not at all reliable to grade a player, let alone 100-200 for the GCL and Williamsport prospects.
* What’s the frame of reference. What is it that this statistic is measuring exactly? Does it tell me something I can’t see just by eyeballing a stat line?
* Baseball Prospectus has Pecota, an evaluation tool which takes further data into account and also produces projections. Sonar seems like a lesser version of what has already been produced in the statistical community.
Just my thoughts. I intend this as constructive criticism. I think too often there’s a tendency to place too much faith in an uber-statistic just because it is presented as such.
LikeLike
Aligning with consensus?
Logically, it seems that the test of SONAR would be how predictive it is compared to consensus. Don’t see how aligning with consensus has anything to do with it. Also,, while I am hearing here that SONAR does not take projection into account (size, physical abilities, etc.) –except for age–that does not mean that it is simply descriptive and not predictive. Isolating the right performance factors can well be more predictive than some traditional criteria.
LikeLike
PECOTA is very useful for guys with large sample sizes of data in the major leagues, but its not very useful when looking at prospects, IMO. PECOTA cards are only generated for some prospects, Domonic Brown for instance doesn’t have one, and what they’re giving you is translations and estimates.
Look, its easy to just pile on and say that this is garbage. I can see why that would be an easy response. But I’ve gone to painstaking lengths to say that this is going to be a work in progress, its going to have blind spots, and its going to take time to see if its providing valuable information. I’m all for constructive criticism, but it seems like people either aren’t taking the time to understand what its trying to capture or looking at it in the proper context. I’ve never trumpeted this as some kind of replacement for scouting reports or other opinions. In fact I’ve gone out of my way to say how important those things are.
LikeLike
PP: Keep doing what you’re doing, it’s useful and seems like something that is fun for you. I enjoy it, as apparently many other people who read this site do.
You must realize that if you put something together like this and present it to the public, it’s going to be questioned and some people are going to criticize it–especially because, as others have pointed out, the results don’t seem to match up to professional consensus, and no one really knows if you’re actually qualified to put together something like this that deserves to be considered accurate. So don’t take it so personally; again, you had to realize it was going to be picked apart, right?
In the end, nothing is valuable which can’t withstand criticism. So don’t take it personally, keep working at it, put it out there, and if it’s worthwhile and valid, time will prove it so. The quality of the work is ultimately all that will matter.
LikeLike
I’ll give SONAR a few years and see if it proves to be an effective evaluating or predictive. Keep up the great work.
LikeLike
Sonar is a work in progress. Keep up the good work.
Leading with the numbers of players with a positive score,
may not be the best use of Sonar, and might have been best not to lead with it.
Over time, you will find things that help and project success.
Match “conventional wisdom” and find things that “conventional wisdom” miss.
LikeLike
One thing I do not get regarding the SONAR scores for hitting: Jiwan James had a positive score, while Anthony Gose had a negative score, but you rated Gose ahead of James in your own “hitting tiers” column. Seems to contradict the stat.
LikeLike
For me, its to early to tell if the Sonar scores are a good evaluation tool or not. I’m generally skeptical but am willing to give it some time before making a final determination.
LikeLike
I too will say, James ignore the non-thoughtful/unhelpful critiques. Knuckleheads abound.
Personally this topic got me to make my first post in over a year of avidly following the site so I have found it to be very fun to think and talk about.
In terms of evaluating SONAR’s success or failure, I do not see how it is possible to even guess at such an evaluation for at least another 3-4 years. Only at that time, at the earliest, will we know if players have repeated high scores or one year outliers and how those repeated high scores correlate to MLB success.
At this point what I think is so fun and potentially helpful to James as he continues to refine his tool are patterns like PhillyFriar’s observation of ARL, or hypothesis on why the non-consensus players/teams may actually be true and the consensus wrong.
For those pining for the formula, you simply do not need to know what is involved in it because you know what it is supposed to do: measure the actual performance of players over their minor league career based on peripheral data. Like stocks, the theory is that although past performance does not guarantee future performance, they are surely closely connected. So regardless of what data James used to come up with that performance number, we can participate in evaluating (especial over the next few years) whether or not his number really did indicate a player’s eventual major league success.
@Allantown I think I understand your confusion. I think you are confusing high top 10 SONAR scores with players advanced in Minor League systems. The SONAR score as I understand it does not strongly take into account success at A- vs AAA. So young low level players can have similar scores to older high level players assuming they perform equally well in the various areas. Thus, we have the result for the Phillies where there are still quite a few players who performed very well last year (high top 10 Sonar score) however they are mostly young players at very low levels (chaos in voter rankings).
LikeLike
PRD says:
“I too will say, James ignore the non-thoughtful/unhelpful critiques. Knuckleheads abound”
hey, i think its good that most people don’t seem to trust this system. we don’t know if its any good yet because its only been in exsistance for like 6 months, and we don’t know how its calulated. its better to be doubtful than to blindly trust something just because we like the guy who made it up.
not all critisim is going to be helpful, some is just going to bash it. and you know what, I don’t think that that’s a bad thing. for God’s sake, this is the internet. Phuture phillies is maybe the nicest, cleanest forum on the whole internet, and everytime people start to get a little bit chippy and maybe say something bad about the sites founder they get blasted to hell.
deal with it, its the internet, you posted something online, its gonna get picked apart.
also, nobody is insulting it, just nobody trusts it because we don’t even know if it makes sense.
LikeLike
Why do people insist that every comment made to the management is an insult. PP has my regards and full backing but he isn’t the Pope. Relax everyone the days are getting longer and spring is coming.
To tell the truth I check the time of sunset and sunrise everyday. Now that is sick.
LikeLike
My subjective rankings of prospects doesn’t contradict an objective stat. SONAR is a mechanical output based on raw data. My personal rankings and thoughts on prospects might differ. I ranked Gose ahead of Jiwan James in the tiers (and in my personal top 30) because hes performed in a full season league and he’s younger than James. SONAR is a data point, its not the be all end all when it comes to evaluating prospects. I’ve spelled that out like 15 times now, and it seems people still keep skipping right over it.
LikeLike
-jpd
“hey, i think its good that most people don’t seem to trust this system” I completely agree. Look elsewhere (the sonar tab section) and you’ll see I have my own serious questions and doubts.
“some is just going to bash it. and you know what I don’t think that that’s a bad thing.” yeah clearly I get that too, which is why I say “knuckleheads abound” and don’t bother engaging such people (unless they engage me specifically) and encourage PP to just ignore them as well. Although, not to get philosophical I would say bashing things just out of habit or because you can is bad.
“and everytime people start to get a little bit chippy and maybe say something bad about the sites founder they get blasted to hell.” I call people knuckleheads or get “chippy” as you put it and you blast me. I pretty sure that whatever you accused me of your doing the same thing.
“also, nobody is insulting it, just nobody trusts it because we don’t even know if it makes sense.” Sid Beam says “Sonar is crappa” That is clearly an insult so your just wrong here.
“deal with it, its the internet, you posted something online, its gonna get picked apart” I think its clear I’m ok with that too.
LikeLike
My own take on the SONAR scores is that it does a good job of highlighting some players whose raw numbers are deceptive, for reasons of age/level or context.
I tend to think that if they are going to be more than that, the system needs refinement. But even as they are, they have value – if treated as “a data point, its not the be all end all when it comes to evaluating prospects.”
Query why the Phillies are so low in terms of number of positive sonar scores? Does it have something to do with drafting/promotion patterns? That first chart doesn’t seem to have much if any coorolation to quality of farm system (while the others do).
LikeLike
The formula needs to be tweaked in order to be taken seriously. The Phillies have a 24 year old catcher in A ball, with 75 abs, a .204 avg, a bad K% and a .092 iso. He still has a positive SONAR score.
LikeLike
Wow guys just pile on!
Look PP is using this as a statistic – in no way is he saying that this is how we grade out prospects and the guy with the highest Sonar score is the best prospect – it’s just one more TOOL to help us see how these guys stack up.
PP isn’t taking into account past performance, signing bonus, scouting reports, or draft round – so YES some guys that scouts, fans etc rate very highly based on tools, talent, etc are not going to have a high Sonar score. PP isn’t saying that this makes them a bad or lesser prospect. I’m sure when we see PP’s top 30 he’ll have guys with low Sonar scores ranked above guys with high Sonar scores.
Sonar is just a way to look at one years performance. PP certainly weighs other factors when considering prospects for himself.
LikeLike
Hey anonymous. The Phillies do not have a catcher who meets that description. After looking around my rosters I figured you meant Brandan Akashian. They do not have him, they released him in the offseason, so it should say they had a Catcher last season, who met that standard.
He had a good rating based on last season’s stats, according to the formula. based on having the highest OBP of all Phillies farmhands. Like said, this formula does not take into account factors other than statistical performance in some instances, and OBP might be an unassailable be-all-and-end-all shibboleth, by some who began talking up this stat at a time when some in the mainstream media did to a much lesser extent. Therefore some may have developed a feeling of ownership of this stat. In some ways it may be over-rated in certain instances.
LikeLike
No matter how hard you tweak a formula you’re still going to have fluke cases like Akashian. The only good way to fix that is probably to require a 150 PA minimum before you’ll rate a prospect.
Keep in mind though that if a guy is reaching base 44% of the time in his brief appearances, he is more than likely to be a net positive for his team performance wise, no matter what the other numbers say.
LikeLike
To me, SONAR has begun to accomplish what it’s inventor set out to do, and this is give us another data point to help identify players in the low minors that others have overlooked based upon performance.
If I were to equate this to baseball, I would say the process is in short-season ball and showing some promise. There are obviously some instruction (tweeks) that can still be worked into the system (player) before it goes big-time (majors). But, I can certainly see some potential for stardom.
To those who won’t believe in this until you see the formula, I can understand where you are coming from. But in this world of copyrights, patents, and attorneys, do you really expect somebody to just give away his secrets that may (or may not) be profitable in the future? That some can’t understand that is just dumbfounding.
LikeLike
And having a “positive score” doesn’t tell you all that much, because a score of 1.00 is still a well below average prospect. This posting, just the raw numbers of positive scores, wasn’t meant to be all that decisive, it was simply something I passed along as I started putting together the individual position rankings. The real meat and potatoes will be the top 30’s, which will combine both hitters and pitchers. Really old guys who dominated rookie ball will have positive scores, but they won’t have very high scores, and they won’t crack their team’s top 30. Akashian is a great example. His numbers were good, his score was positive, but his score wasn’t very high, and he isn’t a real prospect.
LikeLike
“Keep in mind that if a guy is reaching base 44% of the time, he is more than likely a net positive for his team.”
Not in this case.
In this case, you have a player that is much too old for his level, standing at the plate, with no intention of swinging and making minimal contact when he does swing. High BB%, high K%, low average, ridiculously low Slg%. His positive score is not an indicator of his skill. It is an indicator of inexperienced A ball pitchers who can not throw strikes to a player who has no business at that level. Something is wrong with a data point that rates this guy’s hitting above Anthony Gose, an 18 year old who held his own in A ball.
LikeLike
Anonymous. I’m not talking about Akashian’s future prospects with that statement. I’m just dealing with the here and now. If a player produces a quirky line like that, he’s an asset to his team. The problem of course is that the performance is unsustainable. But it explains the positive rating in this case.
LikeLike
I think the problem that I have, is that this Data point needs caveats associated with each player to explain the number associated with his season. The explanation of age to level and number of ABs are part of the stat itself. There should not have to be a caveat that explains that this guy is 24 and this guy is18, therefore the 24 year old’s number is not really an indicator that he is a better hitter than the 18 year old. Those caveats are already contained within the stat.
Gose has been proclaimed an above average hitter due to age relative to level. Yet his SONAR score makes him a non-prospect.
LikeLike
Not sure I understand your point Anonymous. How is a player’s Sonar score any different than any other measure? Stats like .ISO, .OBP, BA, Slg%, etc. are also just a simple statistical measure that doesn’t account for Age, level, etc.
Does Michael Taylor’s .227 batting average he posted in low-a ball mean he wasn’t a prospect or does it simply show that he didn’t have a good year in short-season ball?
I look at the Sonar # just like you would any other stat, some players are going to put up numbers over short periods that are outliers to the measure. Have enough of those outliers and the statistic starts becoming suspect.
But in the end, what this shows is why advanced metrics will never replace scouting. Statistical measures can be a helpful tool but they never tell the full story.
LikeLike
But you’re missing the point “Anonymous”
This isn’t about using one piece of data and saying “so and so is a prospect” or “so and so isnt a prospect”, this is using a data point and saying, what does this tell me about player X? In Gose’s case, everyone knows that he has amazing raw tools, but his approach at the plate leaves a lot to be desired. Getting on base is the core of every offensive player’s approach. If Gose can’t get on base either by hitting for a high average or drawing a lot of walks, then his upside as a player is limited. SONAR sees his deficiencies and points them out in the form of a low score. That doesn’t mean he isn’t a prospect, it means he has a lot of work to do.
This is the disconnect between raw tools/hype and performance. Gose may well be a monster in 2010, draw tons of walks, and do lots of amazing things. But he hasn’t done it yet. His stolen bases are indication of his raw speed and his upside potential. But he’s still very raw, which means we need to be cautious on him.
At the same time, a player may have a high SONAR score and it could simply be a case of him being very old for his level. The point of the entire thing is to provide you with a data point for further investigation, or confirm/deny a thought you had about a player and lead you down a different road. There is no one single statistic that can tell you everything you need to know about a player.
LikeLike
About Akashian…although it may seem quirky to have a score like that look at his numbers from past places he played before this season.
2006 Holy Cross (NCAA Div I) 54 ABs .464 OBP .333 AVG
2007 Holy Cross (NCAA Div I) 106 ABs .520 OBP .330 AVG
2008 Holy Cross (NCAA Div I) 153 ABs .450 OBP .320 AVG
2008 So. Illinois (Frontier) 205 ABs .442 OBP .302 AVG
LikeLike