What's new
Fantasy Football - Footballguys Forums

This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

Interview with David and Joe Posted (1 Viewer)

"From Mohawk: Why is there no post-season analysis of how well each FBG prognosticator did?

DD: Mostly because it's somewhat pointless to go backwards. Fantasy football is looking forward by nature. I do a thorough analysis on how well I predicted a team's run/pass prediction each week and also for the whole season. The players…well they are the noise elements. Guys emerge, others get injured, etc. It's also extremely difficult even coming up with a system that accurately rates how well you predicted things. Let's say you had a QB high and he was the best in the NFL after 10 weeks and then missed the last 6 weeks to finish as QB14. Should you really be penalized that you had him as QB6 higher than everyone else? We do these internal things you are stating here, but we use those findings, etc to get better data going forward.

JB: We do much of this informally and internally and we control who does our rankings and how they're presented. We know that every staffer represents Footballguys and it's up to us to "protect the Black Eyed Joe" to borrow a phrase from Roger Goodell. That to me means making certain I'm comfortable with every staffer's ability to rank and forecast players."

I completely disagree with this. I will say it doesn't need to be done for each individual staffer, but I think it should be done for the VBD cheatsheet and the DD app.

I want to be able to see where I made good/bad/ugly decisions in my draft and who I could have/should have taken instead.

Being able to plug your scoring system into the VBD or DD after the season is over and have it give you a printout of where every guy actually ended up would be very useful. Obviously we would all be able to see how injuries, etc affected it, but it would still be very useful.

It would also allow us to plug in scoring systems for leagues we may be joining in the future and see where those guys ended up under that scoring system for the year.

I don't really care about each individual staffer....I want to know how good the VBD app and the DD were.

good stuff in the interview, thanks for taking the time....

 
Last edited by a moderator:
Great read.

I figure this is probably pretty impossible to do right now, maybe even for a long while but what about allowing users to create there own chat rooms about specific topics within the framework of fantasy football? I understand that there may still be quite a bit of traffic but it would add another dynamic to the board and allow users to participate with each other in real time and develop better relationships on the board and would also allow topics to be debated on a wider spectrum.

 
Thanks for doing this. While I disagree about the end-of-year analysis of your projections (especially since you point out that you do well in the FSTA evaluations), I appreciate hearing your opinions on these topics (especially that you watch every play of every NFL game).

Couple of thoughts/questions:

1. I am looking forward to todayinfantasy.com. Sounds like a dream come true.

2. I appreciate that football is harder to predict than baseball because of small sample size, but every site doing projections has to deal with the same problem, so it's not like you're unique.

3. Where can we get the FSTA evaluation of projections?

 
Every year I'm asked for a subscriber only forum where people essentially pay to have access to a limited forum and every year I say no. I think we'll always have the forums free.
Paying to post = lolGood call.
 
"From Mohawk: Why is there no post-season analysis of how well each FBG prognosticator did?

DD: Mostly because it's somewhat pointless to go backwards. Fantasy football is looking forward by nature. I do a thorough analysis on how well I predicted a team's run/pass prediction each week and also for the whole season. The players…well they are the noise elements. Guys emerge, others get injured, etc. It's also extremely difficult even coming up with a system that accurately rates how well you predicted things. Let's say you had a QB high and he was the best in the NFL after 10 weeks and then missed the last 6 weeks to finish as QB14. Should you really be penalized that you had him as QB6 higher than everyone else? We do these internal things you are stating here, but we use those findings, etc to get better data going forward.

JB: We do much of this informally and internally and we control who does our rankings and how they're presented. We know that every staffer represents Footballguys and it's up to us to "protect the Black Eyed Joe" to borrow a phrase from Roger Goodell. That to me means making certain I'm comfortable with every staffer's ability to rank and forecast players."
I appreciate the answers. However I do not think it is pointless to look backward. Sometimes you have see where you came from before you recognize where you are headed. You say you do that sort of analysis internally and informally anyway so you must see some use for the effort. If someone is good at predicting, I as your customer would appreciate knowing who that is. I pay for a service and it would be great to know who is best at providing it.

I love the site. I love where its going. I do not mean to pick a fight. I'm just saying, you know!

 
"From Mohawk: Why is there no post-season analysis of how well each FBG prognosticator did?

DD: Mostly because it's somewhat pointless to go backwards. Fantasy football is looking forward by nature. I do a thorough analysis on how well I predicted a team's run/pass prediction each week and also for the whole season. The players…well they are the noise elements. Guys emerge, others get injured, etc. It's also extremely difficult even coming up with a system that accurately rates how well you predicted things. Let's say you had a QB high and he was the best in the NFL after 10 weeks and then missed the last 6 weeks to finish as QB14. Should you really be penalized that you had him as QB6 higher than everyone else? We do these internal things you are stating here, but we use those findings, etc to get better data going forward.

JB: We do much of this informally and internally and we control who does our rankings and how they're presented. We know that every staffer represents Footballguys and it's up to us to "protect the Black Eyed Joe" to borrow a phrase from Roger Goodell. That to me means making certain I'm comfortable with every staffer's ability to rank and forecast players."

***

Couldn't disagree with this more. I think that looking back and learning from your mistakes is absolutely crucial in order to be better moving ahead. I know you say that it happens internally, but if it's not shared with the FBG community, how can we learn from it? I do appreciate that you guys took the time to answer these questions but I'm surprised about this response. Just my two cents.

 
It's not worth the time/money to go back and analyze their own rankings. Nothing wrong with that though. If people cancelled memberships over "accuracy of projections" maybe they'd look at it but with all due respect I don't think "accuracy" is what they're selling.

 
At its essence, Fantasy Football is basically about predicting player performance.
I disagree. FF at its essence is about entertainment. I'm sure there are like 50 posters in the Pool who could come in here and talk stats and projections and make it look like FF really is about the stats- but Footballguys has 40,000 subscribers- now besides the 100 people who double subscribe so they get 2 entries into the subscriber contest- most of these people aren't hardcore doing hundreds of hours of research FF players. Most of them are casual- want to read about their favorite players/teams and are looking to fill a couple of hours a week (or more) this way. I bet if you took a great tool like rate my team 90%+ of people who entered their teams did so primarily to see 1. which of their teams rated the best2. which football guys liked their team the best3. so they could read the positive comments about the team and pat themselves on the back.Hardcore fantasy guys will look at the weaknesses in their schedule section and fine tune their teams but I would doubt most players pay much attention to this valuable feature. Footballguys is great at entertainment for sure. You have plenty of new stuff up weekly and tons of toys to play with, kudos for your work, but even poker sites like 2+2- which has a ton of professional poker players talking poker- has had its strategy sections dominated by its Other Other Topics sections. Drama, flame wars and Brandi's suicide generated way more traffic than actionable, profitable poker advice. The nature of the beast.So keep up the good work on both the entertainment and information fronts.
 
TS Garp said:
"From Mohawk: Why is there no post-season analysis of how well each FBG prognosticator did?DD: Mostly because it's somewhat pointless to go backwards. Fantasy football is looking forward by nature. I do a thorough analysis on how well I predicted a team's run/pass prediction each week and also for the whole season. The players…well they are the noise elements. Guys emerge, others get injured, etc. It's also extremely difficult even coming up with a system that accurately rates how well you predicted things. Let's say you had a QB high and he was the best in the NFL after 10 weeks and then missed the last 6 weeks to finish as QB14. Should you really be penalized that you had him as QB6 higher than everyone else? We do these internal things you are stating here, but we use those findings, etc to get better data going forward. JB: We do much of this informally and internally and we control who does our rankings and how they're presented. We know that every staffer represents Footballguys and it's up to us to "protect the Black Eyed Joe" to borrow a phrase from Roger Goodell. That to me means making certain I'm comfortable with every staffer's ability to rank and forecast players."***Couldn't disagree with this more. I think that looking back and learning from your mistakes is absolutely crucial in order to be better moving ahead. I know you say that it happens internally, but if it's not shared with the FBG community, how can we learn from it? I do appreciate that you guys took the time to answer these questions but I'm surprised about this response. Just my two cents.
They do look back, they just do not publicize it. And if they did, they key would be how do they fare RELATIVE to other sites (not internal to this one) which we already have some data on.
 
I think that one thing that could be useful in publishing information would be to compare the past performances of David Dodds, Bob Henry, Maurile Tremblay, and Jason Wood, but just amongst those four. The reason I say that it is because they are the four FBG's who have their projections in the Projections Dominator. Currently in the PD I have all four of them equally weighted when I use it. But if I knew that one of them was more accurate than the other three, then I may adjust the weightings in the PD to favor that one in comparison to the others. Instead of 25% across the board, I may put one at 40% and the remainder at 20%.

 
I think that one thing that could be useful in publishing information would be to compare the past performances of David Dodds, Bob Henry, Maurile Tremblay, and Jason Wood, but just amongst those four. The reason I say that it is because they are the four FBG's who have their projections in the Projections Dominator. Currently in the PD I have all four of them equally weighted when I use it. But if I knew that one of them was more accurate than the other three, then I may adjust the weightings in the PD to favor that one in comparison to the others. Instead of 25% across the board, I may put one at 40% and the remainder at 20%.
They are most likely worried about ruining the credibility of their staff.If we all start doing as you said and lowering certain staff members rankings because of an end of year results report then there would be no point in having certain staff members still on board.So you go from having a crew that everyone loves and respects to having 1 or 2 guys holding the fate of the ship in their hands.Personally I'd love to have that tool available from a selfish standpoint so I could pinpoint the person that is actually the best of the best, but I also understand that it doesn't make any sense for them to do it as a company.However, it would've been nice if they had given a more realistic answer than the cop-out approach that they used of it being simply pointless to look backwards.
 
"From Mohawk: Why is there no post-season analysis of how well each FBG prognosticator did?

DD: Mostly because it's somewhat pointless to go backwards. Fantasy football is looking forward by nature. I do a thorough analysis on how well I predicted a team's run/pass prediction each week and also for the whole season. The players…well they are the noise elements. Guys emerge, others get injured, etc. It's also extremely difficult even coming up with a system that accurately rates how well you predicted things. Let's say you had a QB high and he was the best in the NFL after 10 weeks and then missed the last 6 weeks to finish as QB14. Should you really be penalized that you had him as QB6 higher than everyone else? We do these internal things you are stating here, but we use those findings, etc to get better data going forward.

JB: We do much of this informally and internally and we control who does our rankings and how they're presented. We know that every staffer represents Footballguys and it's up to us to "protect the Black Eyed Joe" to borrow a phrase from Roger Goodell. That to me means making certain I'm comfortable with every staffer's ability to rank and forecast players."
I appreciate the answers. However I do not think it is pointless to look backward. Sometimes you have see where you came from before you recognize where you are headed. You say you do that sort of analysis internally and informally anyway so you must see some use for the effort. If someone is good at predicting, I as your customer would appreciate knowing who that is. I pay for a service and it would be great to know who is best at providing it.

I love the site. I love where its going. I do not mean to pick a fight. I'm just saying, you know!
Completely agree. As someone who had 2 prior careers in prognostication (actuary, stock analyst), if you are going to throw a bunch of projections out there and claim that they mean something, then it is important to incorporate the rigor of assessing your past performance.
 
seems that a few people in here haven't seen David's more elaborate response in the other thread.

Sara Holliday (FF Librarian) started doing rankings analysis for all the FF websites via the FSTA. She blogs about it here:

http://www.fflibrarian.com/2010/01/2009-ac...lts-are-in.html

48 sites gave her rankings. We finished 11th.

Donnie from a website called themostcredible (that link now brings back a webpage that my virus protector does not like) presented with Sara and judged 22 different sets of projections. I submitted my numbers and I took second. I don't have a link to share because he gave a pitch at the FSTA and they never posted the results anywhere.

I paid money to Andy Hicks to track weekly projections from all of the pay sites, some free sites, etc. I came in first and Bloom came in second. But the second we even tried to publish the results we had sites telling us that they would sue us, etc. That's the big problem. Competing sites that wing it will never give anyone permission to show results that make them look bad. and even though the content is useless going forward to anyone, it is paid content. We don't have the rights to publish it. Even getting everyone on board on how to judge stuff is next to impossible.

There was another site a few years ago who did this for inseason and we came out on top then too. They quit after a year as a pay website.

Even if no one here believes these results, ask these things from the sites that do weekly projections.

- Add up all the players. What are the total yards, yards per carry, TDs that are expected. Do these conform with league norms?

- How often are the numbers updated after Friday? Many don't change them on the weekend. It doesn't take a genius to know sites predicting 50 yards from the guy that is inactive are going to suck (because others should be higher while he is at zero)

- Do they predict fractional TDs? I say this because it's ridiculous to project either 0 or 1 TDs for 90% of the TEs as most score 3-4 TDs per year.

- How many players are projected for at each position? Some of these sites list 50 WRs. I think we all know that more than 50 WRs will catch a pass each week. In some dual flex leagues, you might need to go 100+ deep just to properly evaluate all of your options.

I KNOW our stuff checks out because of the processes we use. And although Sigmund and I can differ wildly as to the numbers we think overall on a player for a given week, we use similar processes to create our numbers . So it makes sense that in objective studies, we crush this stuff. The other site that always does good in these things is 4for4.com. That's because like us, they simulate games and then make the numbers match this expected result. They also are one of the few websites (besides ourselves) that use fractional TDs. Most everyone else sucks at this unfortunately. I wish I could prove it more easily as it would clearly help sell subscriptions.

I would also ask this question. We have 40,000 subscribers and have grown our numbers every single year. If we sucked at projecting, do you think this would be the case? We have multiple people on staff with statistic backgrounds including a PHD in Stats in Doug Drinen. Do you think he would let us publish numbers that don't conform to historical norms?

I am done with this subject even though it will get asked 10,000 more times before the preseason ends. I get it. People are results driven. They want us to prove to them how good we are. Like someone else said though, the real thing is to show we suck way less than everyone else. But we can't show other people's work without getting sued. and therein lies the world of fantasy information services all claiming they are better than the other guy.
 
seems that a few people in here haven't seen David's more elaborate response in the other thread.

Sara Holliday (FF Librarian) started doing rankings analysis for all the FF websites via the FSTA. She blogs about it here:

http://www.fflibrarian.com/2010/01/2009-ac...lts-are-in.html

48 sites gave her rankings. We finished 11th.

Donnie from a website called themostcredible (that link now brings back a webpage that my virus protector does not like) presented with Sara and judged 22 different sets of projections. I submitted my numbers and I took second. I don't have a link to share because he gave a pitch at the FSTA and they never posted the results anywhere.

I paid money to Andy Hicks to track weekly projections from all of the pay sites, some free sites, etc. I came in first and Bloom came in second. But the second we even tried to publish the results we had sites telling us that they would sue us, etc. That's the big problem. Competing sites that wing it will never give anyone permission to show results that make them look bad. and even though the content is useless going forward to anyone, it is paid content. We don't have the rights to publish it. Even getting everyone on board on how to judge stuff is next to impossible.

There was another site a few years ago who did this for inseason and we came out on top then too. They quit after a year as a pay website.

Even if no one here believes these results, ask these things from the sites that do weekly projections.

- Add up all the players. What are the total yards, yards per carry, TDs that are expected. Do these conform with league norms?

- How often are the numbers updated after Friday? Many don't change them on the weekend. It doesn't take a genius to know sites predicting 50 yards from the guy that is inactive are going to suck (because others should be higher while he is at zero)

- Do they predict fractional TDs? I say this because it's ridiculous to project either 0 or 1 TDs for 90% of the TEs as most score 3-4 TDs per year.

- How many players are projected for at each position? Some of these sites list 50 WRs. I think we all know that more than 50 WRs will catch a pass each week. In some dual flex leagues, you might need to go 100+ deep just to properly evaluate all of your options.

I KNOW our stuff checks out because of the processes we use. And although Sigmund and I can differ wildly as to the numbers we think overall on a player for a given week, we use similar processes to create our numbers . So it makes sense that in objective studies, we crush this stuff. The other site that always does good in these things is 4for4.com. That's because like us, they simulate games and then make the numbers match this expected result. They also are one of the few websites (besides ourselves) that use fractional TDs. Most everyone else sucks at this unfortunately. I wish I could prove it more easily as it would clearly help sell subscriptions.

I would also ask this question. We have 40,000 subscribers and have grown our numbers every single year. If we sucked at projecting, do you think this would be the case? We have multiple people on staff with statistic backgrounds including a PHD in Stats in Doug Drinen. Do you think he would let us publish numbers that don't conform to historical norms?

I am done with this subject even though it will get asked 10,000 more times before the preseason ends. I get it. People are results driven. They want us to prove to them how good we are. Like someone else said though, the real thing is to show we suck way less than everyone else. But we can't show other people's work without getting sued. and therein lies the world of fantasy information services all claiming they are better than the other guy.
wow....the first sentence of that last paragraph seems like this is a touchy subject....call me crazy, but if it will get asked 10,000 more times beore the season starts, that might be a sign of something to take a look at :thumbup: ....that's 25% of your subscribers right there....and maybe something that deserved a better initial answer than "it's pointless to go backwards"......obviously this is something that the subscribers would like to see, but FBG has no intention of providing......I don't think anybody is real concerned about how FBG compares to other websites, but would just like to see some in-house stats....many are either committed to this site already or not (paying subscribers).......

i guess for most of us that use a league managment service, you can always just look at the end of the year results, by position, on your league website (rankings) and compare it to your pre-draft cheatsheet from FBG's VBD or DD, or the individual staff projections and see how they stacked up.....

bottomline....even though they won't give it to us, it would be great to see the individual staff members break down their projections and see where they hit and missed....what they might have overlooked.....or what they might have put too much emphasis on....why things worked out the way they thought or why they didn't....most of the time these guys probably get it right, but nobody is perfect or else this hobby would be pretty boring.....theres nothing wrong with totally missing on something or having just a bad year all around, it happens....those confident in there abilities can probably handle it.....good and bad........

 
Good stuff, but I also disagree with it being pointless to go back. We shouldn't go back merely for the sake of going back but there's certainly stuff we can learn.

What if you saw one staffer had a consistently high hit ratio? Wouldn't you want to pick their brain to see if they were merely lucky vs. being onto something when it comes to predictions?

It seems you're not interested in going back as it would show this game is mostly luck and the big thing you bill your company on (predictions/rankings) is a big pile of no value for the customer.

 

Users who are viewing this thread

Back
Top