Welcome to Our Forums. Once you've registered and logged in, you're primed to talk football, among other topics, with the sharpest and most experienced fantasy players on the internet.
He did a bad job. He used a situation not analogous to ours, at all. It was so simple that it was flat out wrong. Because his scenario was not at all like the one in which we are in, it's ramifications are not applicable.Uniqueness is vastly overrated.
Agree. That example would make sense if the rules were that everyone with one same guy that scores the same gets eliminated, even if 90% have him, but that's not how this contest works. If 90% of the remaining players had the same guy and he has an off week, a good portion of those 90% will still make it to the next week, depending on who else they have, whereas if that guy has a big week, most of the other "unique" 10% will be gone. The rules of this contest are very different from the rules in that example.
No it's not. This is a terrible example that has no parallel to this contestwell putI'll make it simple. Say you have a totally random coin-flip contest between you and 9 other people. You all choose heads or tails and whoever chooses wrong gets eliminated. The winners flip again and start the process over. When the coin is in the air, you disclose which side of the coin you chose. Which of these would you rather have be the case?...Scenario 1:You - HeadsOther guys - Tails, Tails, Tails, Tails, Tails, Tails, Tails, Tails, TailsScenario 2: You - HeadsOther guys - Tails, Heads, Heads, Heads, Heads, Heads, Heads, Heads, Heads?Of course, you'd rather be the one guy whose choice was unique. Under scenario 1, you have a 50% shot at winning the contest on the first flip and 50% chance to lose. Under scenario 2, you have a 50% chance of losing on the first flip, but if you win the coin toss, you still only have a 11% chance to win the contest.Not a perfect example, but you should get the idea by now. Assuming that all else is equal (and you think your players are as good as everyone else thinks their players are), uniqueness is very very good.With that said, all that really matters is that your players are better than everyone else's players. But when the probability of you winning this contest is very very small, you want to have players that the other contestants don't have.'Ahmad Rashad said:Question for those who emphasize uniqueness in this contest:If I'm completely unique with 3 of the 6 most-owned PK's and 2 of the 6 most-owned D's, why do I need to worry about uniqueness for the rest of my roster?It seems like it would be pretty darn hard to come up with a roster that's not unique, so I don't get why we need to strive for uniqueness in putting together rosters. I would think it's much more important to score higher, but maybe I'm missing something...
Already having regrets about my lineup as that scrub WR that I got rid of last night was Ogletree.
Of course, another half like that and he will have almost exceeded 3 out of 4 of his projections for the entire year which are shown on the website.
I looked many times to try and squeeze 3 bucks from somewhere, but NO, could not find a way. Idiot. I know it's 1 half of the 1st game, but he looks like he was worth 3 bucks so far
Studs > dudsThanks for the 1st page update.Shocked at the overwhelmingly large amount of people using 18 roster spots
I didn't see this before I wrote my earlier post, this is an excellent example of the potential benefit of uniqueness.In the coin flip example, you could be the easy winner, or you could be the only guy out in round 1. I like it.I'll make it simple. Say you have a totally random coin-flip contest between you and 9 other people. You all choose heads or tails and whoever chooses wrong gets eliminated. The winners flip again and start the process over. When the coin is in the air, you disclose which side of the coin you chose. Which of these would you rather have be the case?...Scenario 1:You - HeadsOther guys - Tails, Tails, Tails, Tails, Tails, Tails, Tails, Tails, TailsScenario 2: You - HeadsOther guys - Tails, Heads, Heads, Heads, Heads, Heads, Heads, Heads, Heads?Of course, you'd rather be the one guy whose choice was unique. Under scenario 1, you have a 50% shot at winning the contest on the first flip and 50% chance to lose. Under scenario 2, you have a 50% chance of losing on the first flip, but if you win the coin toss, you still only have a 11% chance to win the contest.Not a perfect example, but you should get the idea by now. Assuming that all else is equal (and you think your players are as good as everyone else thinks their players are), uniqueness is very very good.With that said, all that really matters is that your players are better than everyone else's players. But when the probability of you winning this contest is very very small, you want to have players that the other contestants don't have.'Ahmad Rashad said:Question for those who emphasize uniqueness in this contest:If I'm completely unique with 3 of the 6 most-owned PK's and 2 of the 6 most-owned D's, why do I need to worry about uniqueness for the rest of my roster?It seems like it would be pretty darn hard to come up with a roster that's not unique, so I don't get why we need to strive for uniqueness in putting together rosters. I would think it's much more important to score higher, but maybe I'm missing something...
i don't think so. kind of.your example is great if this were a 1 week contest, or a total points over the season league but it's not. For the first 13 weeks, it's better to be an average team every week as opposed to the best team for 12 weeks and the worst team for 1.i would say uniqueness is a bad thing for the 1st 13 weeks. uniqueness is a good thing for the last 3 weeks.I didn't see this before I wrote my earlier post, this is an excellent example of the potential benefit of uniqueness.In the coin flip example, you could be the easy winner, or you could be the only guy out in round 1. I like it.I'll make it simple. Say you have a totally random coin-flip contest between you and 9 other people. You all choose heads or tails and whoever chooses wrong gets eliminated. The winners flip again and start the process over. When the coin is in the air, you disclose which side of the coin you chose. Which of these would you rather have be the case?...Scenario 1:You - HeadsOther guys - Tails, Tails, Tails, Tails, Tails, Tails, Tails, Tails, TailsScenario 2: You - HeadsOther guys - Tails, Heads, Heads, Heads, Heads, Heads, Heads, Heads, Heads?Of course, you'd rather be the one guy whose choice was unique. Under scenario 1, you have a 50% shot at winning the contest on the first flip and 50% chance to lose. Under scenario 2, you have a 50% chance of losing on the first flip, but if you win the coin toss, you still only have a 11% chance to win the contest.Not a perfect example, but you should get the idea by now. Assuming that all else is equal (and you think your players are as good as everyone else thinks their players are), uniqueness is very very good.With that said, all that really matters is that your players are better than everyone else's players. But when the probability of you winning this contest is very very small, you want to have players that the other contestants don't have.'Ahmad Rashad said:Question for those who emphasize uniqueness in this contest:If I'm completely unique with 3 of the 6 most-owned PK's and 2 of the 6 most-owned D's, why do I need to worry about uniqueness for the rest of my roster?It seems like it would be pretty darn hard to come up with a roster that's not unique, so I don't get why we need to strive for uniqueness in putting together rosters. I would think it's much more important to score higher, but maybe I'm missing something...
I predict that all 387 of them will survive this week.
From the rules: "Week 1 trim to 13,500 teams (all advance if we have less entries)"Total entries = 13318.
Nice writeup, our philosophies are pretty similar.
Regarding the Steeler RBs, it was an intriguing consideration to lock up 2/3s of their backfield for $8.But there were reasons to take 1 and not the other...Take Dwyer if you want the chance Mendy's knee doesn't hold up or he's not effective running on it, ditto with Redman and his injuries, making Dwyer their RB1 for big chunks of this season.Take Mendy if you like the odds his knee does hold up and he is RB1 for Pittsburgh for the last 12+ games of the season.Taking both covers both scenarios, but at the expense of one of the 2 definitely not contributing points to your contest team.As with just about every roster decision in this contest, it comes down to risk/reward, plus the cap space and position slots you have available.
In general I agree with the bolded. Not that uniqueness is categorically bad for the first 13 weeks, but in general it does expose you more when the simple goal is to make the cut each week.i don't think so. kind of.your example is great if this were a 1 week contest, or a total points over the season league but it's not. For the first 13 weeks, it's better to be an average team every week as opposed to the best team for 12 weeks and the worst team for 1.I didn't see this before I wrote my earlier post, this is an excellent example of the potential benefit of uniqueness.In the coin flip example, you could be the easy winner, or you could be the only guy out in round 1. I like it.I'll make it simple. Say you have a totally random coin-flip contest between you and 9 other people. You all choose heads or tails and whoever chooses wrong gets eliminated. The winners flip again and start the process over. When the coin is in the air, you disclose which side of the coin you chose. Which of these would you rather have be the case?...Scenario 1:'Ahmad Rashad said:Question for those who emphasize uniqueness in this contest:
If I'm completely unique with 3 of the 6 most-owned PK's and 2 of the 6 most-owned D's, why do I need to worry about uniqueness for the rest of my roster?
It seems like it would be pretty darn hard to come up with a roster that's not unique, so I don't get why we need to strive for uniqueness in putting together rosters. I would think it's much more important to score higher, but maybe I'm missing something...
You - Heads
Other guys - Tails, Tails, Tails, Tails, Tails, Tails, Tails, Tails, Tails
Scenario 2:
You - Heads
Other guys - Tails, Heads, Heads, Heads, Heads, Heads, Heads, Heads, Heads
?
Of course, you'd rather be the one guy whose choice was unique. Under scenario 1, you have a 50% shot at winning the contest on the first flip and 50% chance to lose. Under scenario 2, you have a 50% chance of losing on the first flip, but if you win the coin toss, you still only have a 11% chance to win the contest.
Not a perfect example, but you should get the idea by now. Assuming that all else is equal (and you think your players are as good as everyone else thinks their players are), uniqueness is very very good.
With that said, all that really matters is that your players are better than everyone else's players. But when the probability of you winning this contest is very very small, you want to have players that the other contestants don't have.
i would say uniqueness is a bad thing for the 1st 13 weeks. uniqueness is a good thing for the last 3 weeks.
Entirely fair points, and you are completely correct on the opportunity cost of the risk/reward you see elsewhere in the roster choices. I feel that people tend to "put blinders on" across positions to some extent. ie: "Here are my RB byes, they're balanced. Oops, I forgot to look at the combination of WR + RB byes" sort of thing. I know I do this, and I get the sense from many posts here that others do the same. That said, that's entirely what I did when evaluating the Mendy/Dwyer picks... If either one becomes the main RB in PIT, I think we all believe they are both vastly more talented than $4 (or even $8). My bet was simply that I wanted the PIT backfield for $8 versus any other RB I could pick up for that price. Obviously I'm discounting Redman, but I simply don't see the talent with him. But your point is well taken - saving $4 on one of the runners plus one of my 4 defenses, and suddenly I can afford one of the better backup QBs I would have liked to have - and that's simply a risk/reward calculation I didn't make because of my "RB blinders" when making the evaluation. Thanks for the helpful post.Edit:Not arguing the point one way or the other, but I found this interesting:
Total teams with Mendenhall: 2,426Total teams with Dwyer: 1,637Total teams with both: 338Agree that uniqueness is "bad" in the regular season (weeks 1-13), and can be "good" in the playoffs, but I'm not sure it's very easy to pick a unique player that will explode in the playoffs that's still unique by then (uniqueness for a player that outproduces will decrease as his owners survive), and far outscores players for similar cost on other teams.If Benson goes for 30 in the playoffs, you'd still better have him for the $3 it would have cost. Having a pretty unique player go for 30 is great if he's far outscoring other players for the same cost, but if it's $23 Philip Rivers (0.7% owned) while $19 Matt Ryan goes for 40 (34.8% owned), you're behind those Matt Ryan owners, who have 10 extra points and $4 extra for other players. The Ryan/Benson owner would be way ahead of the Rivers owner in this example, with $1 to spare.On the other hand, if that 30 comes from $7 Emmanuel Sanders (1.9% owned), and other receivers score much less than 30 that week, you have a big advantage. However, if others used that $7 on Blackmon (28.1% owned), and he also scores 30, that advantage is gone (unless you also have and used Blackmon's 30). There's a good reason $7 Blackmon is more commonly owned than $7 Sanders, as most entrants think Blackmon is likely to score more than Sanders. I tend to agree with the owners of Blackmon and most of the higher-owned players, so finding unique players that will far outproduce the heavily-owned players for the same price is not very easy.So bottom line is I agree that uniqueness can be good in the playoffs if you have someone who will outscore others at the same price, but I think that will be rare, and it's more likely that uniqueness will help get you eliminated in the first 13 weeks (e.g., Rivers' week 7 bye while more common Rodgers/Brees/Brady/Stafford go off).Thanks for the explanations.
I think the recent big thread about Dwyer is what spiked interest in people taking a flyer on him in this contest.I really wanted Mendy but I didn't like the uncertainty of how his knee will hold up, so ultimately I didn't keep him.On the plus side, I think it's a contract year for him too.
But this works both ways and that's the point. Say your team is the one with e unique RB in your example. If that RB is the one that gets hurt or suffers a bad 3 week stretch, then to the other 249 teams, it might not seem like a big deal. BUT it is a big deal to you, which is what you care about, your team. Now if that other RB has his normally solid stretch of 3 games then your in a hole to 249 teams because you we're unique.'Das Boot said:
'flc735 said:i don't think so. kind of.your example is great if this were a 1 week contest, or a total points over the season league but it's not. For the first 13 weeks, it's better to be an average team every week as opposed to the best team for 12 weeks and the worst team for 1.i would say uniqueness is a bad thing for the 1st 13 weeks. uniqueness is a good thing for the last 3 weeks.'Das Boot said:I didn't see this before I wrote my earlier post, this is an excellent example of the potential benefit of uniqueness.In the coin flip example, you could be the easy winner, or you could be the only guy out in round 1. I like it.'Warrior said:I'll make it simple. Say you have a totally random coin-flip contest between you and 9 other people. You all choose heads or tails and whoever chooses wrong gets eliminated. The winners flip again and start the process over. When the coin is in the air, you disclose which side of the coin you chose. Which of these would you rather have be the case?...Scenario 1:You - HeadsOther guys - Tails, Tails, Tails, Tails, Tails, Tails, Tails, Tails, TailsScenario 2: You - HeadsOther guys - Tails, Heads, Heads, Heads, Heads, Heads, Heads, Heads, Heads?Of course, you'd rather be the one guy whose choice was unique. Under scenario 1, you have a 50% shot at winning the contest on the first flip and 50% chance to lose. Under scenario 2, you have a 50% chance of losing on the first flip, but if you win the coin toss, you still only have a 11% chance to win the contest.Not a perfect example, but you should get the idea by now. Assuming that all else is equal (and you think your players are as good as everyone else thinks their players are), uniqueness is very very good.With that said, all that really matters is that your players are better than everyone else's players. But when the probability of you winning this contest is very very small, you want to have players that the other contestants don't have.'Ahmad Rashad said:Question for those who emphasize uniqueness in this contest:If I'm completely unique with 3 of the 6 most-owned PK's and 2 of the 6 most-owned D's, why do I need to worry about uniqueness for the rest of my roster?It seems like it would be pretty darn hard to come up with a roster that's not unique, so I don't get why we need to strive for uniqueness in putting together rosters. I would think it's much more important to score higher, but maybe I'm missing something...
Exactly. The problem with this example is that it's not an all or nothing proposition. Uniqueness, IMO, isn't a good thing or a bad thing. Our teams are unique enough or will become so as you move down to the final 250. there is no need to purposely seek out additional uniqueness. Think about it this way, say you had only one unique player. You'd need him to outperform the common player. Now, add a 2nd unique player, he still needs to outperform the common player because if he doesn't, you lose the edge you gained from the first unique player.
Where you fall on uniqueness partially goes back to your philosophy for this contest. I believe in doing what's best for the first 13 weeks, and taking my chances in the final 3 weeks, so I tend to believe that uniqueness is an unimportant consideration. If you fall on the other side of the former debate, then you may also align with those who believe uniqueness is important.
Great Job Butcher, well done!
Seems nice in theory. But, it only plays out in that scenario - where there is a 50% chance of one player outscoring another by 5 points. If you think there is a 75% chance that Rodgers outscores Brees, then picking Brees to be unique is a losing proposition.The goal here is to pick the team that will score the most points. Trying to be "unique" will have a much higher fail rate than trying to pick popular players. Sure, Ogletree might blow up and be the #1 WR this year, and the players that picked him will be rewarded - but if they also chose 4 busts, while trying to be unique, they will be punished. The results tell us there are many more busts than top-10 picks. So, to hit on all good plays, and try to be unique, is a bit like finding a needle in the haystack. It can be done, but the odds are stacked against you.
Not sure what you mean it only plays out in that scenario. The whole premise of the argument clearly states that it's only valid IF one specific condition is met, and that condition is that you aren't giving up too much production in order to become unique. Even if you think Rodgers is 75% chance to outscore Brees and everything else in my example remains the same, you're still better off picking Brees. Team's A through D now each have a 19.375% chance of winning, while team E has a 22.5% chance. Of course there is a crossing point where if you lose too much production (in this case, you have too little chance of outscoring common player) that you're better off being common. The point I've stated all along AND am sticking to, is that you don't need to go out of your way to become unique. Each team will become unique enough to win through attrition of the contest. The part of my opinion I've changed is that once in the top 250, I'd rather have a more unique team than a less one.
I agree. I'm not arguing against this. In fact, I've already stated that there is no need to try to be more unique. You're best strategy in this contest is to pick players you think will do the best. Uniqueness will take care of itself.
It's way more complicated than that. And way more complicated than my example. It seems like you're reading unique as picking low cost players that are not owned with a high %. But cost doesn't have anything to do with it. Tom Brady and Drew Brees believe it or not are more unique than Matt Ryan. So I'm not sure I agree that trying to be unique will necessarily lead to a higher chance of your unique player being a bust. Not to mention that my argument is specifically to the Final 250, where if you've made it that far, you've probably hit on most of your unique plays anyway. Plus, even if one of them turns out to be a dud, we haven't even factored (and I'm not exactly sure how you would/could) in the fact that if they are a dud, they don't count toward your score anyway.
I don't like uniqueness for uniqueness's sake, but I definitely believe in having a plausible narrative. I want to look at my team and ask "what has to happen for me to win this thing?". The less complicated and more plausible the answer, the better. That was the philosophy behind my team- lots of commonly owned players (because they're good value), plus a Vick + Stewart kicker. My narrative is "If Vick and Stewart go off, I've got this sewn up". In my opinion, that is a plausible enough narrative (both players have talent and a history of going off), so I'm happy with my squad. I don't think it's very likely, but in a contest of 13,000, even a 0.1% chance to win it all is huge.
I'll start by saying I believe it's probably true that, all else being equal, it's better to be more unique than less unique in the final 250, if your goal is to win the grand prize. This is similar to the idea that small rosters are better because they exhibit greater variance - with a more unique team, you're more likely to end up at the extreme ends of the final standings. You're either going to be near the top, or near the bottom. With a less unique team, you may be more likely to finish in the money, maybe in the top 50, but less likely to be the #1 overall team. I don't know if that's actually true, but it's my intuition. This is where the earlier coin flipping example is somewhat appropriate - if everyone picks heads, they'll all be huddled around each other in the standings, so to speak; if you go against the grain and pick tails, then you're either going to win the thing right away or lose right away. From a contest perspective, if you only care about winning the $20,000, and have no interest in winning $25-50 for coming in the top 100 or whatever, then uniqueness is probably your friend. I think most people probably fall in this boat.
But as with many other contest discussions we've had in the past, we need to remember what we do and don't control. You don't get to pick a new lineup leading in to week 14. You have to make your selections in August. While it's occasionally obvious that some players will have relatively high ownership percentages (Benson this year, Foster a couple years ago, etc.), for the most part you don't actually know how heavily owned a player will be. Furthermore, unlike the hypotheticals that are often presented in this debate, it's never actually the case that 99.9% of the teams own Player A, and you're the only team with Player B, at least not in August when you're making your picks. It's both far more complex and unpredictable than that.
The fact is, as I've said here before, some teams are more unique than others, depending on how you define such a metric, but all teams are sufficiently unique to win the contest. There are no duplicates. There are no rosters that are completely covered by others. You need to put up a string of 200 point weeks in the end - if your players do this, you'll finish near the top. If they don't, you won't. Whether or not they do score that many points for you is independent of how commonly owned they are. Darren Sproles won't be sitting on the sidelines in week 14, checking how many teams still own him in the contest before he decides whether or not to go off for 50 points that week. Besides, if there is a correlation, I'd suspect that the more commonly-owned players are more likely to score a ton of points, since they presumably carried more teams to the final 250, and have therefore been better players during the season. I said uniqueness is good, all else being equal, but all else isn't equal. If Player A wound up on most of the final 250 teams, he must be a pretty good player. He may in fact be more likely to put up a ton of points than Player B who's only on a handful of final 250 rosters - not because he's on more rosters, of course, but simply because he's a better player that year.
Similarly, at the outset of the contest, the most commonly-owned players probably present the greatest value ("wisdom of the crowds" and all that). If everyone identified $3 Benson as the player most likely to outperform his price this year, what does it say about the entries that don't have him? The goal is to load up on as many players that will vastly outperform their price as possible - Benson is the prime example this year, and if you agree, why would you leave him off just to be more "unique?" You'd just be shooting yourself in the foot by intentionally passing on a guy who has the exact attributes you should be looking for.
So when people argue about whether or not uniqueness is good in the final 250, they may be right, but it's sort of uninteresting (to me, at least) because that's something you have pretty much no control over. All of the really good strategy discussion that has taken place over the years deals with the things we do have control over - roster size, budget allocation, bye week strategies, etc. Uniqueness in the final 250 falls pretty far outside that scope. Build your roster with the intention of scoring the most points you can, and let the uniqueness part of it sort itself out.
I think that's a healthy approach, though I'm sure I can look at any entry (well, any reasonably well-constructed entry) and construct a plausible narrative for them to win. If there was some issue with a bunch of entries all being remarkably similar, then uniqueness might play more of a role in determining a winner. But I'm convinced that every entry is sufficiently unique to win, even if it contains many commonly-owned players. The guy who won the $20,000 last year did so in part because Matt Stafford put up ridiculous numbers in weeks 15 and 16, and he was the #1 most commonly-owned player to start the year and was on over 50% of the final 250 rosters. He used Aaron Hernandez all three weeks in the finals, the 15th most popular choice last year and on a third of the final 250 rosters. Roddy White and Calvin Johnson made significant contributions and they were found on 12.4 and 9.2% of the final rosters, respectively. When Antonio Brown went off for 26 points in week 14, 53% of the field owned him. When Lance Moore matched that total in week 15, 35% of the field owned him. Of course, he also got points from lesser-owned players as well, but they didn't count any extra. He won because he put up a cumulative total of over 600 points - 80 of which came from Stafford, 80 from Roddy White, 60 from Hernandez, 60 from Calvin, etc. to go along with the 50 he got from Spiller (1.2% owned) and 35 from Deangelo (3.2% owned). Obviously this is just a single example, but I think it highlights that it's not critical to go out of your way to ensure uniqueness on your team, especially when you can't predict ownership stats anyway. He won, not because he was "more unique" than any other entry in the contest, but because the players he chose happened to put up a bunch of points when he needed them to.
