2024 Projections

[RAY@HQ]

As far as why the projections have changed slightly, I think I’ve already explained that as best I can.

As for Hernaiz, that could happen on the old site as well. If PT gets allocated to a player who doesn’t not have a baseline projection preloaded, then he is all zeroes until I add a projection for him. We reload the projections every day but I do not update those baselines daily. It’s a manual process that I do a couple of times a week at this time of year.

The same goes for the Sean Bouchard example. I had a lengthy discussion about examples like that (I think we were talking about Christian Arroyo) and explained how I handle those examples. Again, these procedures are exactly the same process as the old site.

in general, the point I think you are underestimating is that the projections are never a finished product. They are being iterated and refined regularly, especially between now and Opening Day.

One other point: I make it a point to not look at other projections systems. So I have no insight into any difference between our projections and other sites. But using those other sites to suggest something is wrong or has changed in our system is a logical fallacy.

[Snider]

I realize projections are just projections and in most cases they are going to be wrong by a sizable factor when taking all of the variables into considerations. But they are a supremely important tool for auctions in trying to calibrate how much to bid for a given player. I cannot tell exactly when i signed up to a subscriber of BHQ but I am going to guess it was 20+ years ago when Ron Shandler started the website. I have utilized your projections solely during all of the years and I have come up with dollar values for the players/pitchers based on your projections. I could be way wrong about my perception and if I am please don't hesitant to tell me. But for some reason whether it was the negative AB's in your original projection forecasts or so many players that had no projections next to them in your original work or it seems as if there were so many players that needed to be cleaned up and those projections weren't believable or even today there seems to be more projections that aren't credible - whatever it is i am under the impression that BHQ is putting less emphasis or less passion in providing a one stop shop to utilize projections to come up with a perceived dollar value for our draft day. Maybe it is because much more time and emphasis is being directed at everyone's request to have the functionality of the site improve to what we were used to prior to this big change. I don't know what it is - but because of so many projections that seem questionable I am for the first time ever thinking I cannot use BHQ's projections and am left trying to figure where to go. Just one example - last year there were 12 SP's who had 20 or more QS and 19 SP's who had more than 18. In your projections you have one pitcher Gerrit Cole with as many as 18 projected QS. There are more than a few SP's that have over 100 IP's projected and zero QS. I realize we are still 5-6 weeks from opening day but it seems as if your QS projections are way low even at this juncture in the offseason.

[circle360]

I realize projections are just projections and in most cases they are going to be wrong by a sizable factor when taking all of the variables into considerations. But they are a supremely important tool for auctions in trying to calibrate how much to bid for a given player. I cannot tell exactly when i signed up to a subscriber of BHQ but I am going to guess it was 20+ years ago when Ron Shandler started the website. I have utilized your projections solely during all of the years and I have come up with dollar values for the players/pitchers based on your projections. I could be way wrong about my perception and if I am please don't hesitant to tell me. But for some reason whether it was the negative AB's in your original projection forecasts or so many players that had no projections next to them in your original work or it seems as if there were so many players that needed to be cleaned up and those projections weren't believable or even today there seems to be more projections that aren't credible - whatever it is i am under the impression that BHQ is putting less emphasis or less passion in providing a one stop shop to utilize projections to come up with a perceived dollar value for our draft day. Maybe it is because much more time and emphasis is being directed at everyone's request to have the functionality of the site improve to what we were used to prior to this big change. I don't know what it is - but because of so many projections that seem questionable I am for the first time ever thinking I cannot use BHQ's projections and am left trying to figure where to go. Just one example - last year there were 12 SP's who had 20 or more QS and 19 SP's who had more than 18. In your projections you have one pitcher Gerrit Cole with as many as 18 projected QS. There are more than a few SP's that have over 100 IP's projected and zero QS. I realize we are still 5-6 weeks from opening day but it seems as if your QS projections are way low even at this juncture in the offseason.

Its Feb 19th. Opening day isnt for another 5 weeks. Most people's drafts arent for another month. Most of the teams arent reporting until today or tomorrow. So a lot of stuff is in flux. Just like it was last yr and the last 20 + years at this time. Nothing has changed. Except your perception. Since the site got revamped. Now you seem to be nit picking on everything.

The less passion providing a one stop shop. Cmon now. Yes, they want you to go to other sites to take their business from them. That's like going to Chilli's and the waiter recommending TGIF's to you or something.

[RAY@HQ]

I could be way wrong about my perception and if I am please don't hesitant to tell me. But for some reason whether it was the negative AB's in your original projection forecasts or so many players that had no projections next to them in your original work or it seems as if there were so many players that needed to be cleaned up and those projections weren't believable or even today there seems to be more projections that aren't credible - whatever it is i am under the impression that BHQ is putting less emphasis or less passion in providing a one stop shop to utilize projections to come up with a perceived dollar value for our draft day.

I can't manage your perceptions. Only you can do that. But you're overcomplicating this, and seeing things that aren't there. Yes, we had a projected data set on the old site before we went live at the new site. Yes, the new site projections were kind of a mess, and it took us about 10 days to get that cleaned up. Since the middle of last week when I stated that the projections were back in a good state, you have continued to raise issues with the projections, which of course is fine... but the answers from me have demonstrated, I think pretty conclusively, that the questions you have are not related to the new site. And we're going to do that again today, see below....


There is no strategic or directional change about how we do our projections. However, I do think you have a different objective, or standard for accuracy, in our projections than we do. As I said yesterday, I don't think there's a hill of beans of difference between a projected .395 SLG and a .401 SLG. This all goes back to the other thread about The Great Myth of Projective accuracy discussed in this recent thread. Don't get me wrong, I appreciate that the forum posters pull apart the projections, that's all part of the process that makes them better. But your lens for accuracy seems a little too rigid... today's QS question is a good example, see below...

Maybe it is because much more time and emphasis is being directed at everyone's request to have the functionality of the site improve to what we were used to prior to this big change.

Sorry, this is too funny to just let pass, questioning whether too much of my time is being spent on customer service, in post Kids for Closer Trade of this thread that is virtually your dedicated support line.

I don't know what it is - but because of so many projections that seem questionable I am for the first time ever thinking I cannot use BHQ's projections and am left trying to figure where to go. Just one example - last year there were 12 SP's who had 20 or more QS and 19 SP's who had more than 18. In your projections you have one pitcher Gerrit Cole with as many as 18 projected QS. There are more than a few SP's that have over 100 IP's projected and zero QS. I realize we are still 5-6 weeks from opening day but it seems as if your QS projections are way low even at this juncture in the offseason.

Several of my responses in this thread (and others) have tried to shed light for you on how the projections work. This QS "issue" is another example. There were 1700 QS in MLB last year. Today there are 1634 QS in our projections. The reason the top-end pitchers are slightly lower is that we tend to not project those guys for as many IP as they threw last year. Cole and Webb led MLB with 24 QS last year... that was in 209 and 216 IP, respectively. We have them projected for 196 and 189 IP, respectively. Projections project regression... that's how these things work.

Let me summarize this: It's true that I've spent a lot of time doing customer service over the last two weeks. That was not unexpected. I realize this entire migration was jarring to people, and to you specifically. But this process has been just about exactly what we anticipated here: we expected we would be in a pretty stable place two weeks after the migration. In terms of our data, we absolutely are there. But from here til Opening Day, this is going to be a basically normal spring training for us. And yes, that means iterating the projections... adjusting playing time projections, adding projections for newly-added players like Hernaiz as mentioned yesterday, cleaning up nonsensical projections like Joe Borchard mentioned yesterday, filling in missing QS projections for part-time SP, etc. All of that is exactly the kind of work I do on the projections every February and March, and will be doing again.

Those are the facts from my perspective. But as mentioned up top, I can't manage your perceptions. If the transition of the last two weeks was too much for you to get over, I've already offered a refund to anyone who wants it: just email us at support@baseballhq.com and I'll take care of that for you. Obviously I prefer you stick around, but we really need to get past this notion that the new site has somehow done irreparable harm to the projections. That may have been last week's reality, but it's not this week's reality.

[Snider]

First of all - I am not going anywhere. For as much as it seems I am unhappy with this new site - there is still more good than bad. The only way you are going to get rid of me is if you kick me out.

The table below shows the QS distribution amongst the top 12 pitchers from last year. Let's just say we were both right. The innings pitched projected is 9.1% lower than the actual from 2023. And the QS/IP pitched - my questioning is 27.1% lower (7.8% vs 10.7%) than what actually happened last year. And yes I know it is just Feb.19. But I think my point has some validity to it given than 0.078 vs .107 is a big enough gap to question. i also realize that the number of QS is about the same as last year - my point is about the distribution of how they are spread out by pitcher. Thanks for considering.

One more thing - if it was just innings pitched explanation there should be 227 QS distributed amongst those 12 SP's or an additional 62 which equates to about 5 per pitcher. That's a meaningful difference even in projections land.

image.png

[circle360]

Maybe it is because much more time and emphasis is being directed at everyone's request to have the functionality of the site improve to what we were used to prior to this big change. I don't know what it is - but because of so many projections that seem questionable I am for the first time ever thinking I cannot use BHQ's projections and am left trying to figure where to go.

This is the definition of ironic

Maybe it's because more time is being directed at everyone's request. Followed by ray answering your question and 5 more nit picky questions.

If it was march 15th. id get this more. But on Feb 19th. Not so much When so much is in flux still.

You have said that you have been a sub for so many years. ( BTW. So have I. See june 2003 next to my name . So have countless others here as well )

Have faith in ray and the HQ team.

I understand that people have questions and stuff. but then it reaches a point and it's like cmon already.

[JonE]

Snider, I dug out a projection file from early March 2023. Here are the top ten SPs with projected IP. (Actually, it's 17 because of ties.)

Gerrit Cole: 196

Shane Bieber, Corbin Burnes, Aaron Nola, Sandy Alcantara, Framber Valdez: 189 apiece

Max Fried, Kevin Gausman, Julio Urias, Joe Musgrove, Yu Darvish, Alek Manoah, Dylan Cease, Kyle Wright, Robbie Ray, Miles Mikolas, Logan Gilbert: 181 apiece

HQ projected NO ONE to exceed 196. All of these pitchers were projected to earn between 14 and 18 QS.

Did they think that no one would actually post more than 196 IP or 18 QS? Probably not. But they couldn't, or wouldn't, predict exactly who would do that.

Was their list of 17 pitchers identical to the list of 17 pitchers who posted the top IP? Definitely not.

You are right that there is a notable difference between the predicted 2024 and the actual 2023. But this has nothing to do with the new website, and nothing to do with any change to the BHQ methodology. The imprecision is built in to the projection system. You are entitled not to like it. But if you liked it in 2023 you should like it now.

[Timmeh49]

FWIW, looked my final spreadsheets from the last few years (this would be just prior to my league's auction, which is typically the weekend after opening day) to get some historical perspective:

2023: 2019 QS projected, 1683 actual MLB QS
2022: 1899 QS projected, 1776 actual MLB QS
2021: 2011 QS projected, 1584 actual MLB QS

[circle360]

Snider, I dug out a projection file from early March 2023. Here are the top ten SPs with projected IP. (Actually, it's 17 because of ties.)

Gerrit Cole: 196

Shane Bieber, Corbin Burnes, Aaron Nola, Sandy Alcantara, Framber Valdez: 189 apiece

Max Fried, Kevin Gausman, Julio Urias, Joe Musgrove, Yu Darvish, Alek Manoah, Dylan Cease, Kyle Wright, Robbie Ray, Miles Mikolas, Logan Gilbert: 181 apiece

HQ projected NO ONE to exceed 196. All of these pitchers were projected to earn between 14 and 18 QS.

Did they think that no one would actually post more than 196 IP or 18 QS? Probably not. But they couldn't, or wouldn't, predict exactly who would do that.

Was their list of 17 pitchers identical to the list of 17 pitchers who posted the top IP? Definitely not.

You are right that there is a notable difference between the predicted 2024 and the actual 2023. But this has nothing to do with the new website, and nothing to do with any change to the BHQ methodology. The imprecision is built in to the projection system. You are entitled not to like it. But if you liked it in 2023 you should like it now.

HQ has always been conservative with IP projections.

I have a pile of old Forecasters. I grabbed a random one earlier today. ( 2018 forecaster ). The highest IP projections were Verlander and Max at 200. They had both thrown well over that in the previous seasons.

[Snider]

Thank you guys for your feedback. I really appreciate your constructive criticism. You guys made excellent points on distinguishing between the new website projections and the old one. And yes I have most definitely scrubbed these projections harder than any of the previous ones because they had a fair amount of issues with them when they were first relaunched two weeks ago. With that said whether it has happened in prior years, those projections are good to compare to but for now I am more interested in the 2024 projections. I acknowledge the lower IP that BHQ projects versus the actual of some pitchers that throw more than 200 IP. But the table above actually shows pretty conclusively that the ratio of QS/IP for the top pitchers' projections seem to have a systematic bias to the low side. BHQ can utilize whatever projection system that works for them - but I am just pointing out where there might be an improvement or a bit of a disconnect. Ray has answered my questions on whether anything different is happening with the projection system this year versus prior years. And his answer is no so I accept that. You guys have shown some history to the projections that I didn't look at so that is good. But could the QS projection numbers stand to be increased for the top starters above as well as some others - my guess is probably so.