So many fantasy draft questions can be answered by thoroughly examining our league's scoring rules -- and specifically the typical point gaps of each position's elite, near-elite, and next-tiered players generally perform.
Imagine a league where the top QBs, RBs, WRs, TEs, Ks, and DSTs score roughly 300-350 points most years, while the next tier of players at each position usually score around 250-300, and the next tier 200-250. Obviously leagues don't work this way (or at least, no league I've ever seen), but for the sake of simplicity, we'll use it. So when drafting, we'd know with some certainty that there's no advantage to picking one position over another at #1 overall. And in round 2, if there were four QBs, RBs, Ks, and DSTs off the board, but only two WRs and TEs, based on our player projections, we'd probably pounce on a WR or TE if they're more likely to score in the 300-350 range than any other available player at another position.
Obviously, in most leagues these tiers gaps aren't so neat and tidy: ranges generally narrow as players get worse, and each position often has different elite and near-elite ranges, with QBs and RBs usually at the top, and (obviously) kickers and DSTs usually at the bottom.
But this doesn't mean we can't assess player values based on positional scoring gaps. It just takes a little more effort. For example, using ESPN's fairly generic PPR scoring, last year Russell Wilson was clear-and-away the best fantasy QB, outscoring #2 Cam Newton by 48 points. But the difference between Cam and the #13 QB (Blake Bortles) was just shy of 50. The year before, the gap between the #4 and #15 QB was also just under 50 points. The year before, the gap between the #4 and #16 QB was 45 points.
So to summarize, in recent seasons the scoring gap between the best elite QB and the best next-tiered QB (usually the fourth of fifth best overall) has been wider than the gap within the tier of 12-13 QBs falling below the "elite" grouping (around 5-thru-16, give or take a QB). So in a two-QB league, if you drafted the #1 and #15 best QBs, you'd be better off than if you drafted the #5 and #6 best QB. Statistically that makes sense. And yet many people don't think about it when drafting. The fact is, Russell Wilson (#1 QB) and Tyrod Taylor (#16) combined for more points last year than Alex Smith (#4) and Kirk Cousins (#6). By understanding the size and placement of typical scoring gaps, we can make the most of each draft pick.
Things get more interesting when we look at RBs in relation to QBs. Moving down from the #1 RB, the first time we encounter a 50-point production gap across 10+ RBs is at #11 (extended down to #22). In 2016, the first large-group 50-point gap occurs between the 10th and 21st best RBs. Despite the 2015 season resulting in slightly narrower scoring gaps near the top, 2014 and 2013 saw a long drop (#12 RB and #13 RB, respectively) before seeing a 10+ player gap of 50 points or less.
What does this all mean? In four of the past five years, if you had the fifth-highest-scoring fantasy RB on your team, he was more valuable positionally than the fifth-highest-scoring fantasy QB. Why? Because the drop from #5 to #15 has been consistently larger for RBs than for QBs. We know this intuitively. It helps explain why year after year so many RBs come off the board in the first 10-15 draft picks.
This is why you hear me pushing sometimes for an elite QB (like an undervalued Cam Newton), but not for a top 6-8 QB, who usually won't be significantly better than the top 14-16 QBs. But a top 6-8 RB -- if he plays as expected -- will be far more effective than his QB counterparts when considering positional players 8-10 spots below. In other words, I'd rather have the #5 RB and #10 QB on my team than the #5 QB and the #10 RB, purely based on recent scoring trends.
But of course, it doesn't end here. By carefully examining each position's typical tiered scoring breakdowns, we can identify where we're getting the best value.
DSTs are another interesting case. Last year 53 points separated the third- and 18th-best DSTs, while 39 points separated the #1 from the #3. That's not unusual: in recent years it's been fairly typical for the top two or three DSTs to be clear-and-away better than the next 12-15. That helps explain why so many of us stream defenses all season. Having the seventh-best DST is somewhat meaningless, because you might be averaging only one point more per week than someone with the 12th best. So we stream based on matchup. But if you have an elite DST, you're earning two, three, sometimes four points more per game than most opponents' DSTs. Sure, you could still stream, but you've already got a strong enough edge to roll out that DST most weeks.
We devise and implement a draft strategy and roster management approach that works for us. But in the end, fantasy titles are won with points. Even the best game plans fall victim to bad luck and/or bad decisions. We look for edges wherever we can find them. And one important edge is understanding the quantitative basis for our qualitative game plan.
Imagine a league where the top QBs, RBs, WRs, TEs, Ks, and DSTs score roughly 300-350 points most years, while the next tier of players at each position usually score around 250-300, and the next tier 200-250. Obviously leagues don't work this way (or at least, no league I've ever seen), but for the sake of simplicity, we'll use it. So when drafting, we'd know with some certainty that there's no advantage to picking one position over another at #1 overall. And in round 2, if there were four QBs, RBs, Ks, and DSTs off the board, but only two WRs and TEs, based on our player projections, we'd probably pounce on a WR or TE if they're more likely to score in the 300-350 range than any other available player at another position.
Obviously, in most leagues these tiers gaps aren't so neat and tidy: ranges generally narrow as players get worse, and each position often has different elite and near-elite ranges, with QBs and RBs usually at the top, and (obviously) kickers and DSTs usually at the bottom.
But this doesn't mean we can't assess player values based on positional scoring gaps. It just takes a little more effort. For example, using ESPN's fairly generic PPR scoring, last year Russell Wilson was clear-and-away the best fantasy QB, outscoring #2 Cam Newton by 48 points. But the difference between Cam and the #13 QB (Blake Bortles) was just shy of 50. The year before, the gap between the #4 and #15 QB was also just under 50 points. The year before, the gap between the #4 and #16 QB was 45 points.
So to summarize, in recent seasons the scoring gap between the best elite QB and the best next-tiered QB (usually the fourth of fifth best overall) has been wider than the gap within the tier of 12-13 QBs falling below the "elite" grouping (around 5-thru-16, give or take a QB). So in a two-QB league, if you drafted the #1 and #15 best QBs, you'd be better off than if you drafted the #5 and #6 best QB. Statistically that makes sense. And yet many people don't think about it when drafting. The fact is, Russell Wilson (#1 QB) and Tyrod Taylor (#16) combined for more points last year than Alex Smith (#4) and Kirk Cousins (#6). By understanding the size and placement of typical scoring gaps, we can make the most of each draft pick.
Things get more interesting when we look at RBs in relation to QBs. Moving down from the #1 RB, the first time we encounter a 50-point production gap across 10+ RBs is at #11 (extended down to #22). In 2016, the first large-group 50-point gap occurs between the 10th and 21st best RBs. Despite the 2015 season resulting in slightly narrower scoring gaps near the top, 2014 and 2013 saw a long drop (#12 RB and #13 RB, respectively) before seeing a 10+ player gap of 50 points or less.
What does this all mean? In four of the past five years, if you had the fifth-highest-scoring fantasy RB on your team, he was more valuable positionally than the fifth-highest-scoring fantasy QB. Why? Because the drop from #5 to #15 has been consistently larger for RBs than for QBs. We know this intuitively. It helps explain why year after year so many RBs come off the board in the first 10-15 draft picks.
This is why you hear me pushing sometimes for an elite QB (like an undervalued Cam Newton), but not for a top 6-8 QB, who usually won't be significantly better than the top 14-16 QBs. But a top 6-8 RB -- if he plays as expected -- will be far more effective than his QB counterparts when considering positional players 8-10 spots below. In other words, I'd rather have the #5 RB and #10 QB on my team than the #5 QB and the #10 RB, purely based on recent scoring trends.
But of course, it doesn't end here. By carefully examining each position's typical tiered scoring breakdowns, we can identify where we're getting the best value.
DSTs are another interesting case. Last year 53 points separated the third- and 18th-best DSTs, while 39 points separated the #1 from the #3. That's not unusual: in recent years it's been fairly typical for the top two or three DSTs to be clear-and-away better than the next 12-15. That helps explain why so many of us stream defenses all season. Having the seventh-best DST is somewhat meaningless, because you might be averaging only one point more per week than someone with the 12th best. So we stream based on matchup. But if you have an elite DST, you're earning two, three, sometimes four points more per game than most opponents' DSTs. Sure, you could still stream, but you've already got a strong enough edge to roll out that DST most weeks.
We devise and implement a draft strategy and roster management approach that works for us. But in the end, fantasy titles are won with points. Even the best game plans fall victim to bad luck and/or bad decisions. We look for edges wherever we can find them. And one important edge is understanding the quantitative basis for our qualitative game plan.