Writers of Pro Football Prospectus 2008

15 Jul 2009

How accurate is the KUBIAK system?

I've never used the KUBIAK system before, and currently just using it as a tool. How effective has the system been for your performance? How good are the projections? Is it better to use for late round picks and stick to your gut on earlier picks?

I was curious, how do you use the KUBIAK system?

Posted by: dkibiger on 15 Jul 2009

24 replies , Last at 02 Sep 2009, 10:48pm by socctty

1
Re: How accurate is the KUBIAK system?
by JuridianSantaal... :: Mon, 07/20/2009 - 1:32pm

KUBIAK is only so-so at determining differences between similar players. If it has one RB at 11 and another RB at 18, there might be no appreciable difference and its wholly possible that the RB at 18 will outperform the RB at 11.

It is however pretty good at determining players to stay away from

Where it's at its best is in the dramatic stuff that nobody else is seeing or calling. Larry Johnson a few years back, Shaun Alexander after his MVP year (most of the 370 RBs). I use the ESPN Delta that it has a lot and look for people that it has rated dramatically differently. KUBIAK was very high on the Ravens D and Chris Johnson last year, players I got in all of my leagues late because of KUBIAK and were cornerstones to my teams.

I would recommend sticking to your gut particularly in the earlier rounds (It won't help you much if you're deciding between MJD or AP), but just be aware of the players with Red Flags on them and you may want to pass them up. (It has serious reservations about Turner for example, and he is going to be a top 4 pick in most leagues).

2
Re: How accurate is the KUBIAK system?
by dkibiger :: Tue, 07/21/2009 - 6:33am

I saw that about Turner, and was like, "Whoa". What about KUBIAK's thoughts on Julius Jones. That to me seem very bullish. I've been looking at the difference between ADP by position (ranking them) and comparing them to KUBIAK's rank and trying to see where I can get some good value.

Have you ever been burned by KUBIAK? Which players have you seen that look like incredible value this year?

3
Re: How accurate is the KUBIAK system?
by Kulko :: Tue, 07/21/2009 - 1:27pm

Well I have been burned multiple times by sticking too closely to Kubiak when looking for values.

When they go against conventional wisdom, there will be hits and misses mixed up quite evenly so you should always make sure you got a fallback plan.

2 Years ago they were high on Gore, Lewis, Betts and each time I was up one of them was by leaps the best value on the board. So when Gore had goalline issues and Betts vanished behind Portis' career year I had spend my top 3 picks on 2 mediocre RBs, which quite ruined the season.

Last year I changed my approach and it worked much better. I picked Westbrook in round 1 and then filled out my rooster at other positions. Then later in round 6-10 all of Kubiaks Breakout Favorites where still available, and I selected Slaton, Ch. Taylor and Mendenhall in both of my leagues. Of course only one of them turned out to be top tier (Saton), but that was enough to end up 10 points away from our league title (in the other I was stuck with Randy Moss which did not turn out to well)

4
Re: How accurate is the KUBIAK system?
by JuridianSantaal... :: Fri, 07/24/2009 - 1:26pm

Exactly.

Kubiak IMO is at its best for avoiding dangerous players and grabbing players people are down on late. Players like Chris Johnson last year are a good example of how to use it - Find players undervalued by most people and snag them a round or two before they're projected to go. Just because a player is listed as a top 10 RB doesn't mean you have to grab him in that order. If you're the only one who believes in a guy in your draft (like I did with Johnson), you can wait until late rounds where your risk is very mitigated.

8
Re: How accurate is the KUBIAK system?
by Derekcarosi :: Thu, 08/20/2009 - 8:03pm

This is my first time using Kubiak and I like that ADP sheet but where are the breakout favs? Sorry I haven't done a huge amount of work with the workbook yet.

23
Re: How accurate is the KUBIAK system?
by socctty :: Wed, 09/02/2009 - 10:42pm

Look for a player with an ADP Delta rating in the double-digits or higher. Those are generally the players that KUBIAK feels are under-valued according to conventional wisdom.

5
Re: How accurate is the KUBIAK system?
by bmurphyuk@mac.com :: Mon, 08/17/2009 - 7:10am

I find KUBIAK actually helps set my overall strategy...for example, if you know you can get guys like Moreno and Julius Jones in lower rounds, it probably pays to reach for top WRs in early rounds, and maybe somebody like Rivers for good QB value in the 3rd or 4th. I can already predict a lot of my draft in advance, which leaves me with a couple key decisions in early rounds knowing what I can fill out with later...

6
Re: How accurate is the KUBIAK system?
by hoochiemama :: Mon, 08/17/2009 - 7:51pm

Last year Kubiak paid for itself with Chris Johnson and the Ravens D. I also bit the bullet with Mendenhall but Johnson made up for that pick. This year I can tell you that Mendenhall and Julius Jones will wind up on my team. It looks like Hixon and Torry Holt will also.
My feeling is that you dissect the outperform ADP gems and develop a strategy around getting them. I like to do mocks and make sure I know exactly what rounds they all need to be drafted. Like Mendenhall not all will pan out. Which is why you need to find as many of them as possible to increase your odds in the drafts.

7
Re: How accurate is the KUBIAK system?
by Zack (not verified) :: Thu, 08/20/2009 - 5:27pm

Yes KUBIAK talked me into getting Chris Johnson as well. As someone said earlier, it's best at making against the grain picks like Shaun Alexander and Chris Johnson; don't use it for advice on who to take between Steven Jackson and LaDanian Tomlinson.

ESPN's projections are horrid and incredibly conservative. They don't give you any info you couldn't come up with yourself.

9
Re: How accurate is the KUBIAK system?
by billsfan :: Fri, 08/21/2009 - 10:00am

Like any other projection system, it's very hit-and-miss. As many others have said, it hit big on Chris Johnson last year, and was invaluable in not blowing a high pick on Shaun Alexander. On the other hand, it missed just as big on Michael Turner.

In a sense, it's a fairly conservative system. It downgrades many players that end up having perfectly fine seasons, but it also gives you a good feelinng for some late-round steals. If you spend a significant amount of time on this site, or read the book, then you may already have a sense of the players FO like and don't like this year.

More importantly, especially if you are involved in multiple leagues, it greatly simplifies value-based drafting. If you're just in one league, you could cull projections from multiple sources, calculate position baselines yourself, and come up with a good value-based drafting spreadsheet, but when you're involved in multiple leagues of different sizes with different scoring systems (some IDP), and not a seasoned Excel jockey, then it's the sine qua non of preparing for drafts.

(I also like the Eagles)

10
Re: How accurate is the KUBIAK system?
by alsep73 :: Tue, 08/25/2009 - 4:10pm

Is it normal for KUBIAK to make a lot of major changes to the rankings from update to update? I'm in a 10-team league where QB TD's are worth 6 points, so in previous updates it had Brady, Brees and Manning all ranked in the top 12. Since we haven't had our draft yet, I downloaded the latest update, entered in my league's scoring system and was surprised to see all of them dropping to the late teens and early 20s. I keep going over the scoring to make sure I didn't mis-enter a figure, but it all looks right. Now I'm second-guessing my strategy, which was to take a quarterback if I wind up picking near the turn of the first round.

11
Re: How accurate is the KUBIAK system?
by tally :: Tue, 08/25/2009 - 4:40pm

The changes shouldn't be that drastic. It sounds like the QB baseline was somehow altered between entries. I'd look at your entries for roster composition if the scoring section doesn't have an error.

12
Re: How accurate is the KUBIAK system?
by alsep73 :: Tue, 08/25/2009 - 4:49pm

Just did a side-by-side comparison of the Scoring tabs. All numbers match.

13
Re: How accurate is the KUBIAK system?
by tally :: Tue, 08/25/2009 - 4:58pm

Okay, I found the problem.

The Flex baseline in the default setting for KUBIAK plummeted ~50 points, which raised the values of RB/WR significantly. That must be the error that was fixed in Flex leagues.

14
Re: How accurate is the KUBIAK system?
by alsep73 :: Tue, 08/25/2009 - 5:46pm

So how do I fix that? Or should I just stick with the previous update if a new one's not done before my draft?

15
Re: How accurate is the KUBIAK system?
by tally :: Tue, 08/25/2009 - 5:53pm

Does your league use a RB/WR flex position? If so, then presumably the updated version is correct.

16
Re: How accurate is the KUBIAK system?
by alsep73 :: Tue, 08/25/2009 - 6:11pm

Yes, it uses a flex position. So it was the earlier versions that were messed up, and this one is correct? If so, I'll be sliding Brady/Brees/et al down accordingly before we draft.

17
Re: How accurate is the KUBIAK system?
by tally :: Tue, 08/25/2009 - 6:17pm

Yes. From what I can tell, the previous version included TE in the flex, while this one does not, and that changed the baseline.

18
Re: How accurate is the KUBIAK system?
by alsep73 :: Wed, 08/26/2009 - 10:39pm

Okay, so now that they issued yet another update to apologize for yesterday's glitch, it now has Brady, Brees and Manning all back in my top 10, and even higher than they were in the update from four or five days ago. (Brady is the #3 player according to our scoring system.)

So, to sum up, the original rankings were more or less right, there was a temporary glitch that downgraded QBs, that's been fixed, and now Brady's prediction is so high in a league where QB TDs are 6 points that I should think about taking him ahead of anybody but MJD and Peterson?

Sigh... I need my draft to happen already.

19
Re: How accurate is the KUBIAK system?
by ar4t :: Thu, 08/27/2009 - 12:15am

Just relax and use the system to find underrated players and who to avoid. It can also guide you to what your scoring system is making important. I've played in leagues were the defenses show up in the 1st 3 rounds (which is really good to know)

If it's telling you to take Brady and Brees in the top 10 it's probably because your scoring system makes qb have a lot of value compared to other position . Read it as take a good qb early or get hosed in your league.

24
Re: How accurate is the KUBIAK system?
by socctty :: Wed, 09/02/2009 - 10:48pm

This is exactly the way to read it. For shits and giggles, compare a league with 12 teams and no PPR to a league with 10 teams and 1 PPR (like my current league). Vastly different rankings; it shows you what positions are most important to acquire early.

I always break down the ADP rankings to show which positions are going in which rounds. This way, I find that in Round 1, it's RB-centric, and Round 3 is where QBs are going off the table. That shows where the inefficiencies in the convention wisdom are when you compare them to your KUBIAK rankings. Don't view Fitz, Andre Johnson, or Randy Moss showing up as the #1 player as "I better take that particular guy"... view it as "Wow, WRs are really valuable in this league. I better consider taking one early."

22
Re: How accurate is the KUBIAK system?
by johnkorean :: Thu, 08/27/2009 - 12:05pm

Could be that if Hasselbeck used to be a top QB, with him dropping out of the top 10 without anyone else stepping up, the QB baseline went down, so Brady and Brees' FP over baseline went up. I'm in a 2 QB league with TDs worth 6, and the top 4 players according to KUBIAK are Brady, Brees, Manning, and Rivers.

Of course, that's not the way the draft will actually go. Luckily I'm in a pretty permanent league with years of data on the books, so I have historical data to tell me generally how the draft will go (e.g., I can count on an 8th/9th round run on backup QBs). This is my first year with KUBIAK but my 7th year with the league. Combining the KUBIAK projections with my historical expectations of what will be available to me at each pick is what's shaping my draft strategy.

Further, the KUBIAK projections aren't supposed to be a "The draft should go exactly in the order of FP over baseline" model anyway. Otherwise, you'd have people grabbing 6 QBs before their first RB, hypothetically, depending on how the numbers line up. The art of drafting requires long term planning, not blindly following rankings and taking the best player on your board. As a quick, oversimplified point, you must be prepared to answer questions like "Who would you rather have, the number 3 RB and number 4 QB, or the number 1 QB and the number 13 RB?" Quite often, each pick isn't about who's the best right now, but who will be available the next pick around, or 2 rounds from now?

20
Re: How accurate is the KUBIAK system?
by noledan7 :: Thu, 08/27/2009 - 10:18am

Can someone tell me how to update spreadsheet for my league's scoring?

21
Re: How accurate is the KUBIAK system?
by dutchtreat :: Thu, 08/27/2009 - 10:26am

Make sure you are using EXCEL 2003 or earlier on a PC with Macros. MACs won't work.
Go to the Scoring TAB and update all your leagues settings in the white cells. Save and click on generate my kubiak. The Projections tab will then update.

Login or register to post comments