SoxMachine Podcast focusing on Recent James Fegan Story about Advancements.

[Kyyle23]

51 minutes ago, nrockway said:

My thinking is that the sort of math and programming background required to do anything interesting with "big data" specifically is outside the budget and scope of any professional sports team. Like, if you have that skillset, some quantitative trading firm is paying you a lot more than even the Dodgers will. And you probably aren't making slideshows and giving presentations except to other nerds. You would have to hire an entire department of well-paid employees to see any kind of result in a timely manner. at least two phd statisticians, a sql developer, and a fourth person who knows how to talk to people and probably has a minor in physical education. You invest all that and will it even help? The proof is sort of in the pudding when it comes to finance, there's a track record, there's a body of literature, but that doesn't exist for professional sports that I am aware of. It could be groundbreaking.

Also I'm looking at the Dodgers and Rays staff directory of people who work on data analysis, and they're just some dudes. I'm trying to make the point that data analysis and big data analysis are different. The costs are certainly different. You get no argument from me if the Sox just hire some dude to do regular ol data analysis and cook up some hotspot maps. But all that really is is having an enterprise subscription to Statcast and having some person who can, like you say, turn it into a pdf and email it out lol. Boring to me.

it's not being a Luddite, it's a question of "does it even matter if no one can interpret it". or what's the point of a theory with no practical value? lots, I think, just maybe not in this context. or maybe it is. 

I’m not sure why you assume nobody can interpret it or it has no practical value?  The new generation players seem like they are very in tune with it.  I’m not a huge fan of Ian Happ when he calls into the score (because he is just super monotone sleepy), but he talks a lot about the information presented to him and how he uses it both in the box and on the field 

[kitekrazy]

16 hours ago, maxjusttyped said:

Banninster openly acknowledging they have no chance to keep elite talents like Crochet or Cease long-term is pretty bleak.

Other teams do this.  Long term contracts are a higher risk these days with pitchers.  The economics of reality hasn't reached MLB.  Amazing the the salaries of a guy who performs every 5 days compared to a position player. You can move a position player around get make them useful but a pitcher....demoted to the pen making 25M...   

[nrockway]

2 hours ago, southsider2k5 said:

If the assumption is that the people who are creating this data are too stupid to understand it, that's a place I can't go.

That isn’t close to what I said. First of all, they’re not creating the data, the players create the data. They’re analyzing it. And data analysis is hard, expensive, and computationally challenging. And the analysis might be wrong. Most people refer to a giant body of academic literature when deciding which statistical model to use. There’s no such body of work in baseball, it’s uncharted territory. Which I think is really cool but no one is gonna spend money on it. People might work on it for free but they’re not gonna buy the data, if they can even access it. They’ll spend it on finance and real estate, things that are clearly profitable and you need an edge and the statistical edge is quantitatively proven.
 

no one is “stupid” but the models I’m thinking of haven’t been applied to baseball, it will spit out a result, you can interpret it, but will it be truth and will your interpretation be accurate? 
 

I want to see the sport applications but nobody will invest in it because winning isn’t that important or profitable. It might not work. And scouting is probably just a better investment, it’s tried and true and it works. 

[nrockway]

2 hours ago, Kyyle23 said:

I’m not sure why you assume nobody can interpret it or it has no practical value?  The new generation players seem like they are very in tune with it.  I’m not a huge fan of Ian Happ when he calls into the score (because he is just super monotone sleepy), but he talks a lot about the information presented to him and how he uses it both in the box and on the field 

That’s just regular ol analytics. The topic was “big data” and it drew my interest. I’m not saying nobody understands it. I’m saying “big data” doesn’t exist in sports. It could, it might be interesting. 

[caulfield12]

1 hour ago, kitekrazy said:

Other teams do this.  Long term contracts are a higher risk these days with pitchers.  The economics of reality hasn't reached MLB.  Amazing the the salaries of a guy who performs every 5 days compared to a position player. You can move a position player around get make them useful but a pitcher....demoted to the pen making 25M...   

Even John Danks' career falling apart has been used by the organization as justification for no long-term deals....long-term being considered 3 years for Sox pitchers.

That was a decade ago.

[caulfield12]

7 minutes ago, nrockway said:

That isn’t close to what I said. First of all, they’re not creating the data, the players create the data. They’re analyzing it. And data analysis is hard, expensive, and computationally challenging. And the analysis might be wrong. Most people refer to a giant body of academic literature when deciding which statistical model to use. There’s no such body of work in baseball, it’s uncharted territory. Which I think is really cool but no one is gonna spend money on it. People might work on it for free but they’re not gonna buy the data, if they can even access it. They’ll spend it on finance and real estate, things that are clearly profitable and you need an edge and the statistical edge is quantitatively proven.
 

no one is “stupid” but the models I’m thinking of haven’t been applied to baseball, it will spit out a result, you can interpret it, but will it be truth and will your interpretation be accurate? 
 

I want to see the sport applications but nobody will invest in it because winning isn’t that important or profitable. It might not work. And scouting is probably just a better investment, it’s tried and true and it works. 

If you buy some of the claims of DeepSeek (they are alleged) to have run their models off of OpenAI/ChatGPT....then the energy usage, water/electricity consumption and need for high end chips (Nvidia) just fell by a factor of 75-90%.

That will make the analytical models developed by the Cardinals/Astros over a decade ago look like child's play.

 

Is commercial real estate even viable anymore, long-term?

I guess you're referring more to the "ballpark cities" trend of construction around newly-constructed sports stadiums, with the inclusion of retail, condos, restaurants, entertainment zones, bars/clubs/pubs and casinos.

[southsider2k5]

12 minutes ago, nrockway said:

That isn’t close to what I said. First of all, they’re not creating the data, the players create the data. They’re analyzing it. And data analysis is hard, expensive, and computationally challenging. And the analysis might be wrong. Most people refer to a giant body of academic literature when deciding which statistical model to use. There’s no such body of work in baseball, it’s uncharted territory. Which I think is really cool but no one is gonna spend money on it. People might work on it for free but they’re not gonna buy the data, if they can even access it. They’ll spend it on finance and real estate, things that are clearly profitable and you need an edge and the statistical edge is quantitatively proven.
 

no one is “stupid” but the models I’m thinking of haven’t been applied to baseball, it will spit out a result, you can interpret it, but will it be truth and will your interpretation be accurate? 
 

I want to see the sport applications but nobody will invest in it because winning isn’t that important or profitable. It might not work. And scouting is probably just a better investment, it’s tried and true and it works. 

Literally all of baseball is doing this.  

[nrockway]

7 minutes ago, southsider2k5 said:

Literally all of baseball is doing this.  

Go look at the staff directories on mlb.com and check the “data analysis” roles and look at their backgrounds. I only did it for two teams I think have good analytics staff. They don’t have quant backgrounds. They’re dudes like us, frankly. If they’re doing “big data” analysis, I wouldn’t necessarily trust the results. They do it in geography, the field of my interest, and I wouldn’t totally trust the results. I’m interested in the applications, but I think hiring a staff poet might have more impacts on winning baseball games. 

[nrockway]

16 minutes ago, caulfield12 said:

If you buy some of the claims of DeepSeek (they are alleged) to have run their models off of OpenAI/ChatGPT....then the energy usage, water/electricity consumption and need for high end chips (Nvidia) just fell by a factor of 75-90%.

That will make the analytical models developed by the Cardinals/Astros over a decade ago look like child's play.

 

Is commercial real estate even viable anymore, long-term?

I guess you're referring more to the "ballpark cities" trend of construction around newly-constructed sports stadiums, with the inclusion of retail, condos, restaurants, entertainment zones, bars/clubs/pubs and casinos.

Chinese innovation…I won’t touch on that but they seem to be doing good work. Still, I don’t think the Sox are buying a supercomputer. Wouldn’t need it for baseball, it probably wouldn’t actually be computationally challenging, I could’ve left that out. NVIDIA is screwed because all they focus on is video games. Apple is improving in this field, I don’t know how their stock reacted to the nvidia drop. Interesting post, I think you might’ve combined two. 

[caulfield12]

29 minutes ago, nrockway said:

Chinese innovation…I won’t touch on that but they seem to be doing good work. Still, I don’t think the Sox are buying a supercomputer. Wouldn’t need it for baseball, it probably wouldn’t actually be computationally challenging, I could’ve left that out. NVIDIA is screwed because all they focus on is video games. Apple is improving in this field, I don’t know how their stock reacted to the nvidia drop. Interesting post, I think you might’ve combined two. 

Because I tend to look at everything through (first) a stock market/investing lens, and secondarily, societal/sports/White Sox, etc.

I've already learned a HARD lesson here investing in Alibaba (sold 200 Berkshire Hathaway B shares, brilliant) and bought (BABA) when it was down from $340 into the $230's/$240's, Baidu, JD, Ten Cent, etc.   Sold it this past summer at around $105-110, so that was not one of my best investing moments.

Obviously both China and the US are investing billions into this emerging industry, you have Apple (dramatically struggling now over here now, just like Tesla), Meta/FB, Google/Alphabet, Huawei, all the big players involved in both countries....OpenAI potentially for sale (to Musk for around $100 billion?), the overhanging TikTok resolution that's supposed to come in the next 45-50 days or so. 

To be honest, I decided it was much better to shield myself from individual stock volatility and dumb "gut decisions" and just keep most money across an array of mutual funds...basically sold all the individual stocks.

 

 

My dad used to be a big believer in gold as a hedge, but I personally prefer autograph collecting (mostly presidents/celebrities) and have another $50-75,000 worth of baseball cards from the 1950's through the early 90's...and then started getting interested again recently, tried to corner the market on O'Neil Cruz rookie cards, lol.  I bought a bunch of Ohtani, Sasaki and Murakama cards directly from Japan, as well.

Worst/dumbest baseball card mistake was selling a mint 1968 Nolan Ryan rookie card in the early 90's so I could spend the summer traveling and playing baseball instead of working, lol.  The pride of my collection is 2 PSA certified Roberto Clemente autographs and most of his cards from the 50's and 60's, including his rookie card in VG condition.  Sold my 1949 Brooklyn Dodgers' autographed ball as well because nowhere to store my stuff in US anymore.

Edited February 20 by caulfield12

[southsider2k5]

47 minutes ago, nrockway said:

Go look at the staff directories on mlb.com and check the “data analysis” roles and look at their backgrounds. I only did it for two teams I think have good analytics staff. They don’t have quant backgrounds. They’re dudes like us, frankly. If they’re doing “big data” analysis, I wouldn’t necessarily trust the results. They do it in geography, the field of my interest, and I wouldn’t totally trust the results. I’m interested in the applications, but I think hiring a staff poet might have more impacts on winning baseball games. 

There are a ton of assumptions baked in here.

[Kyyle23]

22 minutes ago, caulfield12 said:

Because I tend to look at everything through (first) a stock market/investing lens, and secondarily, societal/sports/White Sox, etc.

I've already learned a HARD lesson here investing in Alibaba (sold 200 Berkshire Hathaway B shares, brilliant) and bought (BABA) when it was down from $340 into the $230's/$240's, Baidu, JD, Ten Cent, etc.   Sold it this past summer at around $105-110, so that was not one of my best investing moments.

Obviously both China and the US are investing billions into this emerging industry, you have Apple (dramatically struggling now over here now, just like Tesla), Meta/FB, Google/Alphabet, Huawei, all the big players involved in both countries....OpenAI potentially for sale (to Musk for around $100 billion?), the overhanging TikTok resolution that's supposed to come in the next 45-50 days or so. 

To be honest, I decided it was much better to shield myself from individual stock volatility and dumb "gut decisions" and just keep most money across an array of mutual funds...basically sold all the individual stocks.

 

 

My dad used to be a big believer in gold as a hedge, but I personally prefer autograph collecting (mostly presidents/celebrities) and have another $50-75,000 worth of baseball cards from the 1950's through the early 90's...and then started getting interested again recently, tried to corner the market on O'Neil Cruz rookie cards, lol.  I bought a bunch of Ohtani, Sasaki and Murakama cards directly from Japan, as well.

Worst/dumbest baseball card mistake was selling a mint 1968 Nolan Ryan rookie card in the early 90's so I could spend the summer traveling and playing baseball instead of working, lol.  The pride of my collection is 2 PSA certified Roberto Clemente autographs and most of his cards from the 50's and 60's, including his rookie card in VG condition.  Sold my 1949 Brooklyn Dodgers' autographed ball as well because nowhere to store my stuff in US anymore.

I genuinely think that while this was a poor future financial decision, this was probably a pretty awesome and memorable summer 

[CaliSoxFanViaSWside]

16 hours ago, maxjusttyped said:

Yup. I've stopped posting on here because constant doomerism & negativity isn't good for anyone. But what's the point of getting emotionally invested in any of the Sox current prospects if you know they'll be gone in 4-5 years if they become stars.

 Max it's never been easy to be a Sox fan.But if it ever turns out that the Sox can become a consistently good team on par with teams that do good work with big data and players development  I'd be thrilled about it. It's not like there's been Sox lifers among the players recently. That era came and went along with the arrival of Free Agency many years ago.

Most fans following the team now were born in the era of Free Agents.

If with a new owner they operate more like Tampa Bay and less like the Sox typified by the Reinsdorf Disinformation Era , I ,for one , would be  happy to root for anyone who plays hard on winning teams.

[WestEddy]

1 hour ago, nrockway said:

That’s just regular ol analytics. The topic was “big data” and it drew my interest. I’m not saying nobody understands it. I’m saying “big data” doesn’t exist in sports. It could, it might be interesting. 

Baseball teams aren't crunching web browsing data from entire populations. Maybe don't get caught up in the semantics of a pun by the article author? Teams crunch data on EVs and barrels based on pitch types and zones, etc. All teams do have their own proprietary models for what they're looking for, and yes, sometimes, they can be wrong. There seemed to be some dramatic disagreement last year between Pedro Grifol and the stats department on defensive positioning, for example. That was probably more on Grifol being a dimwit. 

[GreenSox]

12 minutes ago, CaliSoxFanViaSWside said:

If with a new owner they operate more like Tampa Bay and less like the Sox typified by the Reinsdorf Disinformation Era , I ,for one , would be  happy to root for anyone who plays hard on winning teams.

I'll take operating like the Braves,    Similar budgets/budget ability.*   Braves have just had a well-run front-office.

*At least we have historically; I'm assuming JR will allow a competitive budget.

 

[pcq]

2 hours ago, maxjusttyped said:

Yup. I've stopped posting on here because constant doomerism & negativity isn't good for anyone. But what's the point of getting emotionally invested in any of the Sox current prospects if you know they'll be gone in 4-5 years if they become stars.

Sox fan base on a collective suicide watch. Butterfly collecting would be more interesting. 

[CaliSoxFanViaSWside]

7 minutes ago, GreenSox said:

I'll take operating like the Braves,    Similar budgets/budget ability.*   Braves have just had a well-run front-office.

*At least we have historically; I'm assuming JR will allow a competitive budget.

 

I'm assuming I last a little longer than JR and a competitive budget is realized along with a stronger structure under a new owner.

But you never know, JR could outlast me . 😂

[caulfield12]

57 minutes ago, Kyyle23 said:

I genuinely think that while this was a poor future financial decision, this was probably a pretty awesome and memorable summer 

Yeah... it wasn't so bad, another summer I cornered the market on Benito Santiago rookie cards and dumped them after his long hitting streak.

Experiences/memories >>> money

[chitownsportsfan]

What big data does baseball need? Part of the beauty of the sport is discrete math. We already have linear weights and almost perfectly correct run prediction off event analysis. I would guess even the data on spin rate and location and outcome of every pitch throne the last 5 years can be analyzed using sql. You don't need AI because we have repeated events. Maybe AI could suggest combinations against certain types of hitters but again seems like math that doesn't require AI. Right now AI is good for novel pattern recognition in things like fluid dynamics on Mars or Pluto or diagnosing hard medical problems or creating shortcuts in writing, animation and coding. Things that require orders of magnitude more data points than MLB creates.

How bout the Sox simply start listening to their data guys as a first step!

[chitownsportsfan]

3 hours ago, nrockway said:

Chinese innovation…I won’t touch on that but they seem to be doing good work. Still, I don’t think the Sox are buying a supercomputer. Wouldn’t need it for baseball, it probably wouldn’t actually be computationally challenging, I could’ve left that out. NVIDIA is screwed because all they focus on is video games. Apple is improving in this field, I don’t know how their stock reacted to the nvidia drop. Interesting post, I think you might’ve combined two. 

CUDA bro. Vidya has been an afterthought at NVDA since 2020 with crypto and now AI. They have optimized and created GPUs for LLMs and other big models for years. It is why a 5090 is currently about 2x what it would cost if they were simply catering to gaming.

NVDA has gained all the losses from the Chinese hoax created with (go figure) NVDA chips. Anyways go Sox!