David McWater & His Moneyball-Like Tactics In Measuring Prospects

[Citizen Koba]

I bet you had trouble in algebra & beyond in school.

ANYTHING can be given a value.

This ^. How useful or effective a numerical value is at representing something in the real world is of course the real question, but you can indeed reduce anything to numbers...

I ain't gonna bang on about it but every single one of us is being quantified right now as we operate on the internet, as both consumers and voters having our preferences and opinions logged and numericised to better identify our buying preferences. Do those numbers capture the essence of uniqueness as human beings?

Well, that's an interesting question, but I'd like to hope it doesn't - however they don't need to, they just need to tell the folk who are trying to sell us shit what shit it is that we'll buy (or vote for), just as the metrics used here don't need to capture ever nuance of a fighters character and ringcraft to be able to give a better than random chance of predcting future success.

[HeadBodyBodyBody]

I don't think it's that simple. Certainly one fight wouldn't prove anything towards one guy or the other having the better system. You'd need a 100 fights, a 1000 fights, a million fights to wade through the variance & intangibles. I assume you'd have to look at each system & see what each is looking at & what their fighters have done overall. If you looked at 100 guys by each guy you might start seeing an edge one way or another.

Mostly I just see more info being digested & quantified is gonna equal better odds of success. I think that's undisputable. I mean there is a reason guys who run a bad 40 at the NFL combine drop down in the draft. Numbers matter. Even if they are more subjective numbers. It's not like a guy with slow jab, bad defensive habits & a low ability to be coached is gonna find much dispute among knowledgeable boxing ppl on those realities. Nor will be THAT much dispute on the guy with the great jab, a tight defense & who listens & learns from coaches like God is talking to him. I don't doubt there is a margin of error to be found, but it's not like you can't input that into your formula too. It's hard to argue with math.

And again to be clear these are human beings having a number put on them that's suggesting the best chance of success. It's not guaranteeing anything. A human being can feel the pressure of expectations of potential success n choke. They can break up with their gf & lose focus & desire. Their mom could die n that fooks them up. Not being able to account for everything doesn't make the things you can account for of no value which is what you seem to think.

I think there is some use in it, sure, but this is no paradigm shift. This is still the same old scouting system, with scouts plugging in numbers as they see fit. If this is the case, there is nothing really new going on. In the grand scheme of things, this is a sliver of an edge you are gaining, which in the end relies upon good old qualitative inputs.

This ^. How useful or effective a numerical value is at representing something in the real world is of course the real question, but you can indeed reduce anything to numbers...

I ain't gonna bang on about it but every single one of us is being quantified right now as we operate on the internet, as both consumers and voters having our preferences and opinions logged and nemericised to better identify our buying preferences. Do those numbers capture the essence of uniqueness as human beings?

Well, that's an interesting question, but I'd like to hope it doesn't - however they don't need to, they just need to tell the folk who are trying to sell us shit what shit it is that we'll buy (or vote for), just as the metrics used here don't need to capture ever nuance of a fighters character and ringcraft to be able to give a better than random chance of predcting future success.

I'm surprised at you, Koba, sure h0m0 economicus is a useful model, and economics is the best of all the social sciences, but even this leaves out so much that is important about the human condition. 'GDP measures everything, except that which makes lives worth living.' This is why the search for a typology of extremism is doomed to fail. There is no way to model that effectively. I'd suggest that boxing plays upon similar core truths of what it is to be a human... to love a child unconditionally, to fight for a lost cause, to surrender your soul to the greater good... there's nothing effectively quantifiable about this these sorts of things. We all know that the mental game plays a huge role in boxing, and yet psychology is a very poor science, in terms of predictive power, and all the other traditionally-held parameters of science. Which is what McWater is trying to do here, right, predict the future success of boxing prospects? It's about using the right methods for the right explananda. In physics, it's all about the numbers. In humans, it just doesn't work. We can talk about trends, capacities, averages of averages, but there are no formulas, and a scouting system based upon Google-Analytic-esque data analysis will inevitably miss the really special fighters.

[Citizen Koba]

I think there is some use in it, sure, but this is no paradigm shift. This is still the same old scouting system, with scouts plugging in numbers as they see fit. If this is the case, there is nothing really new going on. In the grand scheme of things, this is a sliver of an edge you are gaining, which in the end relies upon good old qualitative inputs.



I'm surprised at you, Koba, sure h0m0 economicus is a useful model, and economics is the best of all the social sciences, but even this leaves out so much that is important about the human condition. 'GDP measures everything, except that which makes lives worth living.' This is why the search for a typology of extremism is doomed to fail. There is no way to model that effectively. I'd suggest that boxing plays upon similar core truths of what it is to be a human... to love a child unconditionally, to fight for a lost cause, to surrender your soul to the greater good... there's nothing effectively quantifiable about this these sorts of things. We all know that the mental game plays a huge role in boxing, and yet psychology is a very poor science, in terms of predictive power, and all the other traditionally-held parameters of science. Which is what McWater is trying to do here, right, predict the future success of boxing prospects? It's about using the right methods for the right explananda. In physics, it's all about the numbers. In humans, it just doesn't work. We can talk about trends, capacities, averages of averages, but there are no formulas, and a scouting system based upon Google-Analytic-esque data analysis will inevitably miss the really special fighters.

Absolutely man, I wasn't for a second suggesting that we can be summed up by metrics, in fact I thought I'd made that fairly clear in my post - tried to at least. I certainly wasn't expressing approval for the reduction of human behaviour and feelings to numbers. All I'm saying is that for specific narrow purposes, like selling us shit perhaps or maybe for increasing the probability of successfully scouting fighters certain elements of our behaviour and responses can be and are statisically modelled.

You say that a mother's love cannot be expressed in numbers and I wholeheartedly agree but her behaviours can...what % of her disposable does she spend on non-essentials for her child, toys, games, etc? How many minutes will she leave her child crying on the naughty step before compassion overcomes the desire to enforce discipline... these things absolutely are not love, but they are metrics that can be used to quantify certain demonstable behaviours associated with love. That's kinda what I meant when I said anything can have a number put to it. And those kinda numbers are exactly what will be used by adverisers and marketing folk to part that mother from her money. And it works within the narrow confines of what they're trying to achieve.

I don't like it... the idea that aspects of our behaviour can be metricised and modelled can in fact give the dangerous impression that that is all we are, which is absolutely abhorrent to my ethical sensibilities and brings to mind a quote of my user namesake 'one death is tragedy, a million deaths is a statistic'. Seems to me exactly like a recipe for dehumanisation and the reduction of our lives to a role as consumer or commodity... a topic I can go on about at length, though not here perhaps. But it can be done and is done.

In terms of boxing I'm sure you're right, top talents could very easily be missed specially if they're doing something radically unorthodox or different - the example of Dmitry Pirog springs to mind... virtually overlooked as an amateur likely due to his self taught offbeat style not fitting with the (ex) Soviet school, but pretty damn good as a pro...

But.. and this is the kicker, with all this kind of modelling it boils down to probabilities... you might miss some elite level talent but even if you increase your chance of picking out a future champion by 10% or make your product 3% more likely to be bought you're ahead of the loop and people will use this kinda tool. In the case of boxing I personally would be sceptical too... there's just so many different factors involved in the make up of a top flight fighter, but I'd have to imagine the proof of this pudding lies in whether it produces the results it claims.

[Eff Pandas]

I think there is some use in it, sure, but this is no paradigm shift. This is still the same old scouting system, with scouts plugging in numbers as they see fit. If this is the case, there is nothing really new going on. In the grand scheme of things, this is a sliver of an edge you are gaining, which in the end relies upon good old qualitative inputs.

Again bruh my main frustration with discussing this with you is you CLEARLY haven't listened to the podcast or even read what's already been talked about in this thread. There isn't really a scouting system in boxing like in most other sports. That's why this guy might be able to make money at some future date providing his proprietary scouting system to others without taking the risk he's taking now with signing guys.

I listed 29 ppl he's signed since 2016. One not only fought for a title, as this info is trying to predict the chances of a boxer doing, he won a title. If 2 more of these guys fight for a belt he's among the best in the game at scouting talent. If 3 more of them fight for a belt he's the best in the game & clearly onto something.

[Eff Pandas]

I'm surprised at you, Koba, sure h0m0 economicus is a useful model, and economics is the best of all the social sciences, but even this leaves out so much that is important about the human condition. 'GDP measures everything, except that which makes lives worth living.' This is why the search for a typology of extremism is doomed to fail. There is no way to model that effectively. I'd suggest that boxing plays upon similar core truths of what it is to be a human... to love a child unconditionally, to fight for a lost cause, to surrender your soul to the greater good... there's nothing effectively quantifiable about this these sorts of things. We all know that the mental game plays a huge role in boxing, and yet psychology is a very poor science, in terms of predictive power, and all the other traditionally-held parameters of science. Which is what McWater is trying to do here, right, predict the future success of boxing prospects? It's about using the right methods for the right explananda. In physics, it's all about the numbers. In humans, it just doesn't work. We can talk about trends, capacities, averages of averages, but there are no formulas, and a scouting system based upon Google-Analytic-esque data analysis will inevitably miss the really special fighters.

I guess you've never heard of Socioeconomics. Like the idea or not a number can be put on ANYTHING.

It's a fact most people with children are less happy than people without children. That's a measurable fact cuz of Socioeconomics. Does that mean YOU would be happier without children? No. Maybe you would, maybe you wouldn't. There is still variance with numbers that say even though odds are you'd be happier without children maybe that doesn't prove true for you specifically. That's all this guy is doing basically. All you can do is look at the measurable talents & skills that say these attributes equal a successful boxer capable of fighting for a belt more often than guys without them.

And if you listened to the podcast or merely read previously discussed things you'd know this system already has predicted two of the most successful guys of the class of 2016 by saying Teofimo & Shakur are top 4 guys of that class most capable of pro success. Did this data miss there specialness? Doesn't seem so.

[TheMyspaceDayz]

If you believe anything in the human experience can be reduced to numbers then you probably think we are living in a simulation.

[BennyBlanco]

I guess you've never heard of baseball scouts. They assign numbers to attributes. Again I'm surprised ppl are misinterpreting this so badly.

It's a simple matter of measuring more data leads to a more likely to be correct answer to a imperfect question where variance & intangibles are in play.

This isn't really a controversial thing in most other sports. They just got it all aligned a lot better & there are less controversial or subjective ways to measure athletes against each other with things like the NFL scouting combine.

You have it completely backwards.

The Moneyball revolution was a reaction AGAINST the subjectivity of scouts. Moneyball is about objective measurables, advanced metrics, complex statistical analysis and predictive probabilities. Moneyball is the antithesis of a scout eyeballing a prospect and assigning him a number for each of a defined set of qualities.

Yes, you can assign a number to everything - but the Moneyball approach says that the numbers must be objective measurables. A 40-yard dash time, a pitcher's throwing speed, a NFL prospect's bench press, WAR, defensive efficiency all are OBJECTIVE MEASUREMENTS. They don't depend on a scout's opinion.

As long as you are relying on a scout's subjective evaluation of a prospect's characteristics, then it isn't the Moneyball approach.

[Eff Pandas]

You have it completely backwards.

The Moneyball revolution was a reaction AGAINST the subjectivity of scouts. Moneyball is about objective measurables, advanced metrics, complex statistical analysis and predictive probabilities. Moneyball is the antithesis of a scout eyeballing a prospect and assigning him a number for each of a defined set of qualities.

Yes, you can assign a number to everything - but the Moneyball approach says that the numbers must be objective measurables. A 40-yard dash time, a pitcher's throwing speed, a NFL prospect's bench press, WAR, defensive efficiency all are OBJECTIVE MEASUREMENTS. They don't depend on a scout's opinion.

As long as you are relying on a scout's subjective evaluation of a prospect's characteristics, then it isn't the Moneyball approach.

Your not wrong for baseball, but the thing is there is nothing like most pro sports scouting systems in boxing for this system to rebel against in the first place so this system is the reaction to no system.

And I'd argue it's not necessarily a apples to apples comparison to the Moneyball thing in the first place it's just a catchy title ppl have used to put a name on what David is attempting. I believe there are things that are objectively being measured in amateur results as much as they can be measured towards a pro career, but the one issue ppl have brought up that's for legit is there aren't many objective measurements in boxing to begin with. There is no Boxing Combine where all boxers hit a bag to put a number on power, there is no agility test to put a number on that, a reflex test, a wonderlic test, etc & so on to put values on the many attributes that help a boxer to be successful. The mass scouting of boxers IS the Boxing Combine at this point.

Randomly a Boxing Combine would be an amazing advancement in boxing if boxing made any fooking sense. This would allow boxers to auction off their skills & upside right into the pro game to the highest bidder. Only thing with boxing is cuz of guys turning pro all different times of year since boxing has no seasons you'd have to do it monthly or quarterly at least.

[Eff Pandas]

If you believe anything in the human experience can be reduced to numbers then you probably think we are living in a simulation.

You know me better than I'd have thought.

[Eff Pandas]

This ^. How useful or effective a numerical value is at representing something in the real world is of course the real question, but you can indeed reduce anything to numbers...

I ain't gonna bang on about it but every single one of us is being quantified right now as we operate on the internet, as both consumers and voters having our preferences and opinions logged and nemericised to better identify our buying preferences. Do those numbers capture the essence of uniqueness as human beings?

Well, that's an interesting question, but I'd like to hope it doesn't - however they don't need to, they just need to tell the folk who are trying to sell us shit what shit it is that we'll buy (or vote for), just as the metrics used here don't need to capture ever nuance of a fighters character and ringcraft to be able to give a better than random chance of predcting future success.

Exactly. How good or bad this system ends up proving to be is still up for discussion & analysis but I'm completely surprised ppl think more info about a fighter being given with a value to the best of ones ability to assign that value isn't a superior means of accessing talent.