| 0 |
Subject: how much is a trade worth?
Posted by: matic
- [171461913] Fri, Feb 19, 2010, 14:46
I had some free time at work so.... |
| 1 | matic
ID: 351461913
Fri, Feb 19, 2010, 14:46
|
There was a discussion a couple months ago in the TSN boards about the point value of a trade, i think damoose started it. I recognized alot of your names in the discussion. I always wondered the same thing and I had some free time at work so I decided to do a little experiment. Alot of people were saying between 35-45 points and thats what I always assumed a trade was worth as well, but when i crunched numbers I got different result. This doesn't take RV mentality into consideration at all and is just based purely on points.
There's 170 days in the season and divided by # of weeks and the 4 initial trades that gives each manager 100 total trades (correct me if i'm wrong.) If you were to spread the trades out evenly, that means 10 trades per player slot on average. That leads to each player slot having 11 total "player periods" (the ten trades + the one inital draft.) I always thought that each player was on the team for 17 days average but its actually 15.5 days, cut down to 14.9 days average hold (164/11) if u take away ASB/xmas/Easter.
So I took one slot and simulated how I would trade it for the maximum amount of games using the ten trades alloted to that slot, ignoring any moneymaking moves and also assuming no injuries. (Using the trades evenly across ten players most maximizes the amount of gm/trade in the end) I found that through soley trading on schedule linking you can get 105 total games in that slot. Subtracting that from the games a zero trader would play (105-82) you get 23 so for that first slot its safe to say that the optimezed slot got 2.3 extra games per trade. I did the same thing for 2 other slots using different team sequences and got similar numbers 103/104. I know there will be a diminishing return as you add in the rest of the slots as it would be harder and harder to get slots that fit perfectly without overlapping. But in the end I assume you can put the average to about 1.7or 1.8/ games / trade should be what you get. If you wanted to put this into points if you figure your average player yields about 35 a game you can deduce that a trade is worth about 60 tsnp/trade when all is said and done.
I know all that is highly hypothetical, so I did another method using real data. Pretending i'm pro like SU ;), I drafted a pretty awesome team using 50 million (at the players oringal prices.) If I take those players projected total SNP at end of season, using their current average x 80 (i subtracted 2 for injuries and random DNPs) I get about 28000 tsnp. So if i drafted that team and made 0 trades all year I would end with that tsnp. If I compare that to the tsnps of the gurupies top teams over the last 2 seasons (33000-35000) using 34000 as an average There's a 6000 tsnp difference and if divided by the 100 trades they made I find it interesting I end up with 60 tsnp/trade as well.
|
| 2 | money4later
ID: 2111213116
Fri, Feb 19, 2010, 15:33
|
Nice! I always feel like I "wing it" every day of the season, but this is helpful info.
|
| 3 | Headless T Gunners
ID: 38050167
Fri, Feb 19, 2010, 15:42
|
So
If we just picked up Rondo for 2 days
if he doesn't crash and burn tonight...Bingo!
|
| 4 | matic
ID: 521241915
Fri, Feb 19, 2010, 16:24
|
3.
Heh well picking up Rondo for 2 games to fill a 2 game gap with 1 trade seems like its 2 Rondo games / trade but it really isn't because while u do get 2 games out of nothing you are still only holding him for 2 days meaning somewhere in the checks and balances of your team you will have a player that will have to be held 13 days longer (assuming averages) to bring the 2 day Rondo hold into equilibrium.
But I too did pickup Rondo for 2 days to fill my Lopez - Durant hole.
|
| 5 | twolves
ID: 104481913
Fri, Feb 19, 2010, 17:10
|
Interesting analysis. I usually look at trade in hand will allow at least one game advantage over opponent and thus realistically can gain 35-40 points for a "stud" trade at this time of year.
I wonder if a trade is worth more or less to a team with 50 mil RV or 80 mil team if looking at consecutive trades. For one trade probably makes little difference but over time and type of players traded certainly could.
Also it would be interesting to see if the number remains constant for the early part and late parts of the season.
|
| 6 | Guru
ID: 330592710
Fri, Feb 19, 2010, 17:43
|
There will be times that a trade yields a much bigger return. For example, if you have a stud player entering a 1-in-6, and buy a stud with 4-in-6, you can pick up more. But there are also trades that net you less, as with an injury-induced trade.
And there are trades that don't necessarily add points (or many points) in the short term, but that capitalize on favorable price swings. An old rule of thumb that I've used is that a marginal $1 million in gains should generate about 2 points per game over the balance of the season. I haven't reviewed the numbers underlying that assessment lately, but I suspect it is still in the ballpark. So, although a trade might not generate extra points immediately, if it produces a meaningful gain, it can still generate extra points over the remainder of the season.
|
| 7 | matic
ID: 61101919
Fri, Feb 19, 2010, 20:10
|
5 + 6
you guys both bring up good points and Guru u hit it right on the marker with the last statement.
That's kind of what I was trying to point out too, I was just saying that when ALL is said and done you can use the numbers I got to average it out to about 60 a trade and monetary gains could also be averaged to apply to each trade by taking money gained at end of season and divide by 100. Its just to apply an average value to each trade and as you stated yea there are going to be trades that will net you 120 points and 0 $$ gain and maybe the next one will net u 0 points but 1.5 million in cash etc. but all can be averaged to give a value to the single trade at the end.
And yes the trade in the beginning that netted 50 pts and 1 million might be more "valuable" as it effects trickle over the remainder of the season (as guru said) more so then that same trade would have made in April 13th.
Which leads to twolves question, I think its all about relativity to your team in question. The numbers are definitely greater at the end of the season due to sheer amount of RV advantage and will grow more exponentially when we get to all stud teams lineups and probably will be a higher spike for those of us that were able to reach 80 million so fast (not me sigh). But in terms of whats worth more I think its completely even because the trades made at the time of 50 million would have still been relative to the competition level of 50 million RV back then which leads to where we are at today. (From a losing manager's point of view), I'd think its actually more valuable in the beginning, even if the trades actually amount to a fraction of what you can get today, just because its mainly the first two months that determine who will be competitive and who will be be GMD benchwarmers (waves to the top 10).
|
| 8 | Comic Book Guy
ID: 20817160
Sat, Feb 20, 2010, 01:53
|
People sometimes talk about the average value of a trade, but I never think about that. We're going to make 100 trades during the season and will get wildly different expected (as well as actual) results from them. Some trades will be a guy with a bad injury for a star player with a good schedule for a month and yield hundreds of extra points. Some will be a guy who's flaming out for a guy who's on fire and stays hot. Others may just be 2 trades for 2 games of a 40 TSNP player, then back to the guy you already had. I try to do trades like the latter only after the All-Star break when I think I have enough emergency trades because a trade like that can cause you to lose points if you're out of trades before the end of the week and can't do everything you want. Usually if I think I can get 55-60 points on a trade I do it unless it means being tradeless early in the week. Some early season trades are done for money only: a couple years ago I knew Kelenna Azubuike would make a ton of money even though I thought he was pretty dubious. I bought him just hoping he would play decently long enough to make a lot of money. I think I made 800K plus on him and he never made it to the meat part of his schedule.
The point to me is to try to make the 100 best trades I can. Some will be much more productive than others. There is room for a few 35-40 point trades but they should be well-timed.
|
| 9 | Comic Book Guy
ID: 20817160
Sat, Feb 20, 2010, 01:56
|
Didn't want to leave this out: some trades will be utter disasters that cost you points and money.
|
| 10 | reebbertxx
Dude
ID: 561124720
Sat, Feb 20, 2010, 02:10
|
Does 2 trades for 1 Delonte West game equal an "utter disaster"?
|
| 11 | Comic Book Guy
ID: 20817160
Sat, Feb 20, 2010, 02:22
|
To me it counts as two trades only if you went back to the guy you traded for Delonte. I just remember you trading Delonte for a great stretch of Stephen Curry. I've done much worse than use a trade on one sub-20 point game of some scrub. Don't even want to dredge up all the past disasters, but I am in the "I bought Joakim Noah with a bad foot, I'm an idiot" club.
|
| 13 | matic
ID: 21134202
Sat, Feb 20, 2010, 03:34
|
CBG
the points you bring up is one of the reasons why i brought up the average value of a trade. I was trying to point out that people trading out of players in good schedules into other players because of "inklings" as well as trading into no game gain schedules for hopeful RV gains go against the optimal plan. And your'e right sometimes the trade from a "flaming" player to a rising player may yield points regardless of schedule. Like e.g. trading a 20 tsnp 2 mill cheapie for Darren collison over the last month will no doubt yield more then my calculation of 2 gm trade ratio.
I just wanted to make a solid trade/pt ratio that is somewhat accurate based on data. My calculations was just to show that at the end of the season ....gurupie teams can calculate that they progressed on average 60 points per trade used (i used a few different methods to come to this #). If anything the 60 pt/trade conclusion i came to can just be used to assess the trades you made when you are reviewing your team at the end of the season as to whether or not a certain trade you made was the below or above the determined average that the trade alotted.
|
| 14 | Senator Urine
ID: 348262116
Sat, Feb 20, 2010, 09:20
|
That's some really interesting analysis. It's pretty amazing how complicated this game is and how many ways there are to analyze it. A lot of us have been playing for over a decade now and probably haven't even begun to scratch the surface in quantifying things.
"I'd think its actually more valuable in the beginning, even if the trades actually amount to a fraction of what you can get today"
Agree with this completely. There have been a ton of trades I've made at the beginning that in some cases lost me points in the short term, but the money has such a long-lasting effect. Sometimes that trade that loses you 20 points can end up more valuable than a trade that gets you a quick 80.
|
| Rate this thread: | If you wish, you may rate this thread on scale of 1-5. Ratings should indicate how valuable or interesting you believe this thread would be to other users of this forum. A '5' means that this thread is a 'must read'. A '1' means that this is a complete waste of time. If you have previously rated this thread, rating it again will delete your previous rating. If you do not want to rate this thread, but want to see how others have rated it, then click the button without entering a rating, or else click here. |
|
|
Post a reply to this message:
|