This is a follow up to a "Draft 101" post I did awhile back. For the sake of brevity, I am going to assume you read the first post. If not, here is a link: Draft 101
First, thanks everyone for the positive feedback on the original post. The internet can be a harsh critic, so it's great when you give positive feedback. It made my day. Also, thank you Fooch for getting this some extra screen time on the front page. Lastly, thanks to Poldarn for the excellent follow up analysis, and to the many other posters with positive insights to add. As I mentioned, I am no expert. What I am doing, however, is looking at the data and providing the info that seems insightful. I am a data nerd.
The purpose of this post is simply to refine our thinking on draft strategy. With the first post, I really focused on expectations. What can we expect from a specific draft spot and what are realistic trade strategies. The logical extension of that is "how do I use this when doing my mock drafts and should we trade up or trade down?"
Teams don't trade up to get to a draft slot. They trade up to get a player. A prime example is our trade last year for Eric Reid. Our front office negotiated the trade specifics before the draft. They probably called every team and worked out potential deals with at least a half dozen. Something along the lines of, "if we do a deal, this is what we will do." Then, on draft day, both sides either say "let's do it" or "I'll pass" based on how the draft unfolds. Heck, Dallas might have had potential deals worked out with multiple teams, and then we were either the best deal available or the only one that called on draft day.
So, technically we did work out a trade to get to a draft position, but we most likely didn't finalize the trade until draft day, when we saw the player we wanted was still available.
Draft day rolls around and we call Dallas when they are on the clock. What probably made us call? Kenny Vaccaro. Vaccaro was the first safety drafted last year. He actually had a great season for New Orleans but got injured toward the end of the year (Weirdly, he is also a dead ringer for our QB, but that is for a different fanpost). Our FO knew Vaccaro would be the first safety taken, and he went at to NO at 15. We probably didn't have deals, or deals we liked, with the Bills at 16 or the Steelers at 17, so we waited until 18 and then made our move. We didn't think Eric Reid was going to drop to us, so we went up and got him.
Was it the optimal move? We will never know. Once you change one pick in the draft, the "butterfly effect" kicks in. Virtually every pick after the one you change could be different. What I do know is that Cinci, Houston, and the Rams drafted safeties fairly early last year. They were sitting at 21, 27, and 30 in the first round, all in front of the 49ers. Baltimore and St. Louis also drafted safeties at 32 and 33, right after our original pick. So, there was a big demand for safeties. That doesn't mean Reid wouldn't have fell to us at 31.
Ironically, the Bills and Steelers also drafted safeties last year. Baalke just has nerves of steel, and waited it out to 18.
Who "won" this trade? It's another difficult question to answer. Dallas got the center they wanted, and in the process picked up Terrance Williams (WR) with the pick we gave them. Terrance had a pretty good rookie season with 44 catches and 5 TDs. We got a pro bowl level Free Safety. This truly feels like a win-win.
Should you trade up?
The value charts published in the first fanpost and subsequent blog entries analyzed past performance to identify sweet spots in the draft. We saw the places where, if you drafted in that spot every year, over the course of 5-10 years you would clearly come out ahead. The problem is that teams don't draft in the same spot every year, and they don't have full control over where they draft.
We've established that a team trades up to get a specific player they desire. They will use their trade charts to help them determine whether or not they should make that trade.
In our earlier example, Dallas says they want a third round pick to swap firsts with us. We look at our draft charts and say "that 3rd round pick has a 30% chance of being a 3 year starter. Do we want to give that up to make sure we get Eric Reid?" Said slightly differently, we knowingly gave up 30% of a starter to make sure we get Eric Reid.
From a value standpoint, we lost a lot of value to make sure we got our guy.
The rule of thumb is pretty simple, to maximize value never trade up. This is because for a trade up to create value you need to increase the chances of getting a starter more than the chances of the pick you give up. Using the Draftmetrics charts I referenced in the first fanpost, even a 7th rounder has an 8% chance of being a starter. Giving up a 3rd round pick, to get equivalent value, would be like moving from the 3rd to the 1st round. Nobody is taking two 3rds for a 1st. Trades just aren't like that. Most trades are less than a dozen spots, and the opportunity cost of losing an extra pick will always be more than the better draft position you obtained.
(I know there was some question about the draftmetrics data. Poldarn came up with some different numbers, but also used a larger but older data set. We will need to debate this in another post. For those of you statistically inclined, we are using samples of 12-25 with discrete data. Basically, the confidence intervals are huge. Regardless of dataset, the principle still stands.)
We will never trade up again. Now what?
Wait, it's not that simple. If I was running the 49ers draft room, you would never want me to trade up because I don't know what I am doing. The experts we hire, however, have evaluated the talent and have additional information that they believe outweighs the draft averages. Baalke saw something in Eric Reid that was worth 30% of a starter to make sure we got him.
The decision to trade up solely depends on the capability of your front office to effectively weigh the pros and cons of the move and make the correct call. In the case of Eric Reid, our front office hit a home run. Unfortunately, it's not always that pretty. In 2011, here were all the people that teams traded up to get:
|Team||Moved From||Moved To||For||Performance||Picks Gave Up|
|ATL||27||6||Julio Jones||Stud||59, 124, and 2012 1st and 4th round picks|
|JAC||16||10||Blaine Gabbert||Now attending 49ers QB camp in hopes of resurrecting his career||49||Paid Premium|
|CLE||27||21||Phil Taylor||Started 15 games at NT in 2013||70||Paid Premium|
|NO||56||28||Mark Ingram||78 carries in 2013||2012 1st||Paid Premium|
|SF||45||36||Colin Kaepernick||Stud||108, 141||Paid Premium|
|IND||53||49||Ben Ijalana||Tore ACL twice, now on NYJ||152|
|CHI||62||53||Stephen Paea||Started 10 games at NT in 2013||127|
|DET||75||57||Mikel Leshoure||2 carries in 2013||107 plus swapped positions in later draft rounds|
|HOU||73||60||Brandon Harris||Servicable DB, potential starter in 2014?||138|
|MIA||79||62||Daniel Thomas||109 carries in 2013||146, 217|
|JAC||80||76||Will Rackley||IR for parts of 2012 and 2013. Starting LG before injury||182|
|OAK||219||92||Joseph Barksdale||Released in 2012, pickup up by Rams||2012 2nd||Paid Premium|
|BAL||90||85||Jah Reid||Just plead not guilty in battery case (reserve OL)||191|
|TB||116||104||Luke Stocker||28 career catches, none in 2013||2012 4th||Paid Premium|
|WAS||127||105||Roy Helu Jr.||62 carries in 2013||144, 152||Paid Premium|
|ATL||158||145||Jacquizz Rodgers||96 carries in 2013||229|
|CLE||168||150||Jason Pinkston||Started 2011, but injured in 2012 and only active 3 games in 2013||170||Paid Premium|
|NYJ||161||153||Jeremy Kerley||43 catches in 2013||194||Paid Premium|
|SF||174||163||Daniel Kilgore||We shall see||231|
|MIA||179||174||Charles Clay||Listed at FB in 2011. Pretty good 2013 at TE||218||Paid Premium|
I could do something similar with 2012 or 2013, but the results would be inconclusive.
Some of those guys, especially the niners picks, are doing fine, but some of those guys are potentially on their way out of the league. Lots of "meh" in that list, and that is the ultimate challenge of drafting. At the end of the day, it's still gambling. The player you want may end up being great, but he might also get injured or just not materialize at the pro level.
Let's look at the teams that are successful and see if we can find a trend (# trades 2011 to 2013).
|Trade Up||Trade Down||2013 Playoffs|
The team that made the most trades (us) made the playoffs last year and the team that made the least trades (KC) also made the playoffs. There just isn't a relationship that defines good or bad. It's about the players you pick, not how much you move around (duh).
There are some teams that are good at moving up to draft players
Fans assign different start dates for Trent Baalke's management of our draft process. Some people say he wasn't in-charge until 2011. I give him credit for 2010 also, where he took over for Scott McGloughlan prior to the draft. I'm sure Trent had been helping build the draft boards since he joined the organization, and since he made the final decisions in 2010, I am giving him credit.
In you look at Trent's drafts since 2010, the best players he has drafted are:
7 great draft picks, 4 of them in the Top 18 or better. This is not to say that those other 28 draft picks that I didn't list are bad. Some may turn out to be great, but what I do know is that Trent Baalke drafts GREAT when he is working in the top 20. In fact, the only year he has not drafted in the top 20 was in 2012. I don't need to say anything more on that.
Trent has shown a high degree of success with those players that are highly regarded and have a high chance of success (another duh). The 49ers have created the draft equity that allows moving up in the draft, and we have a deep enough roster that we don't need to fill every draft pick we have, so moving up makes sense.
Trading up is not for teams like the Raiders, that need to get as much value as they can out of each draft right now. Luckily for us we are not in that position.
Outside of the first round, our trade results are a little more mixed:
|Traded Up For||Traded Down So Others Could Draft|
|Eric Reid||Justin Hunter|
|Vance McDonald||TY Hilton|
|Corey Lemonier||Lamar Miller|
|Joe Looney||Frank Alexander|
|Colin Kaepernick||Will Rackley|
Overall, it looks like we are winning more than we are losing, but that TY Hilton pass really hurts. All six guys we traded up for could eventually start for us, and we've got a pro bowl safety and a franchise QB mixed in.
How much is a trade up worth?
We see a guy we like that is still available, and we don't think he will fall to our next pick, so we decide to trade up. How much will it cost? We already know we are losing overall value with the trade up, so we want to spend as little as possible.
In the first article, it was pretty clear that teams are using the original Jimmy Johnson draft chart, but are giving discounts. Looking at the results by season:
|Paid Full||Discount||Avg Premium||Avg Discount|
This chart reinforces how poorly team's felt about the 2012 draft class.
Anyway, there are a few patterns in the data, but not much. I think teams come in low, and then end up paying full price in the following scenarios:
1. The move is a large move (e.g. 20 or more spots)
2. Multiple teams are bidding to move up. Happens for players like RG3.
3. In 2011 and 2013, the premiums were paid toward the end of round 1, the beginning of round 2, and the beginning of round 4. The beginning of rounds 2 and 4 represent the beginnings of day 2 and 3 of the draft. Teams probably fall in love with a player overnight and then are more aggressive to start the next day.
For mock draft purposes, blogger nickbradley has suggested using a 10% discount. I think that makes sense, unless you are looking at the beginning of round 2 or 4 or are suggesting a large jump in draft spots. This means that if you are trying to trade up to the 16th draft slot, worth 1000 points, you can probably get there with 900 points worth of picks. If you want to be conservative, go for the full 1000 points.
Teams give a discount because they all recognize that the true value chart is flatter than the traditional trade chart. Basically, lower round picks are worth more than the chart gives them credit for.
Validating our assumptions
We've declared that teams will get the most value from trading down and accumulating picks. The niners have done this, and also accumulated picks through the comp pick process. They aren't complete hoarders, however, as they have traded up to get players more than any other team.
If you are trading down, or otherwise accumulating picks, in theory you should perform better over time. I looked at how many draft picks teams had from 2010 - 2012 and how they have done the last two years. With these small datasets, it is difficult to establish concrete relationships, but I do think the data supports the idea that the best teams have the most draft picks.
|Team||# of Picks 2010-2012||2013 Playoffs||2012 Playoffs|
San Diego seams like the only anomaly in this list. Atlanta made the playoffs in 2012 but their lack of roster depth showed up this year. New Orleans made the playoffs on the strength of a franchise QB, but their roster has been eroding. As we've discussed, having a lot of draft picks is not a guarantee of anything, but you are improving your odds. If San Diego continues to draft guys like Keenan Allen in the 3rd then they won't need a lot of draft picks.
OK nation, anything else we should discuss? Did I miss or butcher something completely? Let me know your thoughts and thanks for reading. Now let's get to the important stuff, CB or WR?