| Football Forum View the Forum Registry |
| |
| 0 |
Subject: Small World Adresses Price Changes Posted by: 6-9 With The Afro - [2245413] Tue, Sep 19, 09:42 Don't know if anyone's seen this yet. (I didn't see anything posted about it) Nothing here that hasn't been figured out by Guru, Richard, Pink Pimp, etc. but interesting to hear them describe it in their own words. Now if they'd only give us their formulas... The Big Question: How Do Our Price Changes Work? By Dean Carrano 9/18/00 What makes Small World games special? A lot of things…but the most obvious one is probably the "market" aspect of our games. You don't merely draft some players in Small World…your squad has a salary cap. For maximum success, you must "play the market" and increase that salary cap, so that you can afford better players. The market is what makes our games uniquely fun and challenging. It also means that you, and owners just like you, control what happens in our games! So, how do you "play the market" and increase your franchise's value? It's easy to see: Spend your money on players whose value will go up. When you sell a player for a higher price than you bought him, you make a profit which you can then use to improve your team. It might help to know exactly how the all-important price changes are determined. Think of every player as being on an elevator which represents his Small World price -- the higher he goes on the elevator, the higher salary he commands. Every time an owner of a Small World team sells the player off his team, the elevator goes down a bit. Every time an owner of a Small World team buys the player onto his team, the elevator goes up a bit. That's the basic idea. Here are the wrinkles. First of all, a sell actually moves the elevator down more than a buy moves the elevator up. That's because every day, new owners join the game and draft new teams. Each new team has a bunch of buys and no sells. So, at the end of the day, our market ends up with more buys than sells. To adjust for that, we make each buy have less "pushing power" than each sell. Second of all, if there is very little trading activity around a player, his price will gradually go down -- just like gravity would bring a real elevator down eventually. We do this because if there is no activity around a player, it means that the market must feel his price is too high. We want to get that player "back in the game", so we gradually move his price down until he reaches the level our owners feel is right. Here's a peek into how we do that. Basically, we ask two questions. First, we ask, "Do a lot of our owners own this player?" Then we ask, "Was this player traded frequently during the last pricing period?" If the answer to both questions is "no", then the player's price drops by a maximum predetermined amount that varies from game to game. If the answer to both questions is "yes", then this aspect of our repricing system does not drop the player's price at all. If the answer to one question is "yes" but the answer to the other is "no", then the player's price drops by an intermediate amount. Let's touch on some other pricing issues that tend to come up. A player's real-life performance has no direct effect on his Small World salary. Only the buys and sells made by owners of Small World teams affect Small World prices! Of course, our owners tend to be very aware of who is doing well and who is doing poorly, and usually buy and sell accordingly. But all is not always equal. If a player's achievements tend to get a lot of publicity -- or, alternatively, if the player tends not to get much attention -- that could affect the tendency of owners to buy the player. Preseason buys and sells do not affect a player's price. In other words, on Opening Day, the player's "elevator" starts on the "floor" that our Pricing Committee has decided is right for that player. This inevitably will have some ramifications later in the season. If a player was a very popular preseason buy, his price has more downside than upside. Plenty of people own him and can potentially sell him, driving down his price. Fewer people can buy him to push his price up, since so many owners already have him on their team. If the player was perceived to be overpriced at the beginning of the year-- and, thus, no one bought him in the preseason -- then his price has more upside than downside. If there continues to be no interest in such a player, "gravity" will eventually bring his price down to a more appropriate level. Some folks have wondered whether the overall price movements are larger on days on which a lot of trading occurs (such as the day after trades are given out.) The answer is "no". If there are many trades on a certain day, each individual trade is worth less. If there are few trades on a certain day, each individual trade is worth more. Either way, the overall market movement is the same. We don't want a player's price to change too drastically as a result of one price change. So we set a maximum amount for each price change. A player's price can't go up or down more than that maximum amount in one price change. The exact amount varies from game to game. If this maximum is reached -- which is a pretty rare thing -- then any trades that would make the price move any further are simply ignored. There is no "carryover" to the next price change. Each game is unique, and price changes may not happen for the same reasons in different games. In any of our games, the largest factor in owners' buying patterns is normally the recent performance of players. But there are always other considerations. For example, in baseball, the biggest price changes have historically occurred when owners rotate between starting pitchers. Owners sell a starter after he pitches and buy another starter scheduled to start the next day. Sometimes, this leads to seemingly counterintuitive results: A pitcher could pitch a terrific game and then go down in price. This makes sense since the underlying strategy is sound. In many sports, owners will buy based on the upcoming schedule. They will sell players who have few scheduled games in the near future, and buy ones who have a busy slate coming up. Again, this may result in price movements that don't reflect the players' recent performance. If you're a little confused, visit our Message Boards to see what owners are thinking when they make trades and remember to check out the articles, posted daily. And remember, every time you buy or sell a player, you put your own two cents in! | ||||||
| 1 | Richard ID: 1719313 Tue, Sep 19, 09:53 |
A great concise explanation of price movements including gravity. I'm glad SW finally has described what we have been guessing at. Richard | |||||
| 2 | Guru ID: 330592710 Tue, Sep 19, 10:21 |
Most of this should be old news to Gurupies, even though it is the first time that some of it has been publicly acknowledged by SW. We've certainly seem ample empirical evidence to have been able to nail down issues like carryover effect, weighting of buys vs. sells, draft impacts, and proportionate (rather than absolute) impact. The most interesting "revelation" is the gravity description. We have never been able to fully reverse engineer the gravity effect, even though we have seen strong evidence that both ownership and recent trading activity play a role. Also, the early football price changes have shown more of a graded gravity effect, which suggests that the approach for this game may be slightly different that in prior games. | |||||
| 3 | HeaTransfer ID: 30619280 Tue, Sep 19, 13:32 | Wow. Great work, Richard, Guru and others who figured this all out a good time ago! | |||||
| 4 | mad scientist ID: 248121010 Tue, Sep 19, 20:12 | Heck, they probably got the ideas on how to do their SW pricing formulas from the Guru and this message forum. Before the invention of 'gravity,' (I think I first heard that lingo here) there was talk on this board that if nobody owns a player, then that player's value should fall until people do buy him. We got tired of seeing people who were hurt/retired/benched with high SW prices. | |||||
| 5 | Power ID: 414352018 Tue, Sep 19, 20:46 | Any thought to posting this highly rated thread to the "KEEPERS"? | |||||
| 6 | Madman ID: 44633210 Tue, Sep 19, 23:02 |
I posted the same revelation on the baseball boards, and missed the fact that you were 12 hours in front of me, Afro! I found this description to be very interesting, but yet unsatisfactory. I hadn't worked with this year's data much, but the notion of an "equalized market movement" very interesting. Does that confirm that they indeed changed their pricing formula back in April? Or were some of those days "unequal"? I remember we discovered the equalization factor for 1999, but the existence of such for 2000 has only been speculation, no? Another thing I was disturbed by was the complacency with which they dealt with the very unintuitive fact that highly drafted players have greater downward price potential. The next step in their economic maturity was to be able to figure out a way to increase prices based on the "flubber" principle (or anti-gravity) rather than strictly using buys and sells. It seems like we have coaxed them so far, yet they still have so far to go . . . | |||||
| 7 | Guru ID: 330592710 Wed, Sep 20, 09:27 |
I'm pretty sure the "equalization factor" has been in place throughout. The early season baseball anomalies were most likely due to some other formula tinkering. My leading guess is that they just adjusted everything by a constant factor - perhaps twice. Don Mathis once offered a more complete explanation of that issue, too. (He even suggested that the explanation wasn't as simple as we were surmising.) But he never delivered. The potential (negative) energy stored in highly drafted players is certainly something that could (and should) be fixed. These players are generally mispriced too low to begin with, and this "feature" only exacerbates the problem for the entire season. Perhaps they could figure out what a draft price change would look like, and then phase it in over the first several price cycles. | |||||
| 8 | Eustacio ID: 51729258 Wed, Sep 20, 11:55 |
I'm not so sure that phasing the draft price changes in over the course of the first few price changes would be the best method to go about taking into account the draft buys, but I do believe that something should be done. Phasing the pricing in would give a false belief to players that people might actually be buying these players during the season, and that in turn would have an affect on the SW Marketplace that shouldn't be there, IMHO. Instead of phasing it in, why not set aside a certain time period (The week before the start of a specific sport) that would act as a seperate market? SW could set the formula so that a player cannot drop below their original price (would prevent any group from trying to lower Player X's price). This would make the draft seem more logical; a draft is all about buying players, not selling them. Every day SW would have their 12:00 repricing, and prices would be set until the next day. This would continue up to the first day of the season. Early researchers would be rewarded. Now those of us who decide to start looking at the game a couple weeks in advance can actually make a slight profit before the season starts by picking up sleepers. Of course, these players still have a potential to drop after the season starts. However, everyone has had an opportunity to make money on these players beforehand, just like in the regular season. This is just a rough idea which would require some fine tuning. Perhaps SW would want to cap the total amount that is possible to gain off of 1 player in that 1 week to a million SWD. | |||||
| 9 | Eustacio ID: 51729258 Wed, Sep 20, 11:59 | Oh, and for those of you keeping score, I work as a Lab Assistant at college and use up to 80 different computers a day, along with my home computer, so my ID will tend to fluctuate a little. I try to stay with my home computer, the 2 "Labbie Only" computers, and the one I use during my programming class, but sometimes another one pops in. This is my (current) home computer ID (already moved a couple times this year). | |||||
| 10 | Guru ID: 330592710 Wed, Sep 20, 13:20 |
Actually, I'm not sure a phase in makes sense either. Why not just have an immediate price change at the beginning of the season based solely on draft activity. For long-season sports like baseball, this would also remove any incentive to wait until after a few games to draft a team. | |||||
| 11 | biliruben ID: 88101720 Wed, Sep 20, 13:28 |
I think a pre-season market, would discourage people from playing. Most draft just a day or two before the season, and it would make folks think they had somehow already fallen behind before a pitch was thrown! I like Guru's idea, just get the re-pricing based on the drafting out of the way on opening day (with caps), get the sleeper prices somewhat up to speed, and then we can play the dang game. | |||||
| 12 | Madman ID: 44633210 Wed, Sep 20, 16:59 |
Although the idea of a one-time price adjustment on day 1 sounds appealing, there are a few negatives: a) shock to old-timers who weren't expecting it, discouraging them right off the bat, b) a direct and sharp inflationary impact. Of course, a) could be eliminated via good communication, and b) could be largely eliminated by a reduction in the "money supply" by simply starting with less money per team. To tell you the truth, however, I don't see what would be terribly wrong with a strict anti-gravity policy. That policy would run something like this: If player X was being purchased (including drafts by new teams) substantially, and if player X was already owned by a high percentage of owners, then increase the price increase for player X by 20K or 30K or whatever. If player X was owned by a high percentage, but being sold at a high rate, do nothing other than the regular price drop. I sincerely doubt that even experienced Gurupies would have a hard time recognizing the differences in pricing policies between the anti-gravity model and others. One other interesting side-effect of anti-gravity is that the season-long downward trend for studs might be mitigated somewhat, for obvious reasons. The only drawback I see is that this is once again inflationary. You are going to increase the prices of popular players, meaning that you increase roster values somewhat. However, once again, you could run a simulation based on trading activity for this last year and figure out how much extra $$$ you are pumping into the system and make counter-balances with the overall $$$$ given out on day 1. To sum up, this method would increase the prices of highly drafted players (who maintained their ownership). In addition, this method would tend to increase upward price movements for extremely popular players, meaning that each SW manager would have to jump on the bargains a bit faster than they otherwise would have. As long as the anti-gravity effect is fairly small, however, I don't think this is a problem (in fact, it could be good). And, for those season-long bargains, this method would at least give their price some float . . . | |||||
| 13 | mad scientist ID: 248121010 Thu, Sep 28, 22:01 |
I'm not sure if this idea has been put forth before, but what about "price-tethering"? We're familiar with gravity (price falls if no-one owns/buys them), and anti-gravity (price holds if there is significant ownership, even if there aren't many buys). (These are my understandings of these 'concepts' and they may not be accurate). What I'm proposing is something similar to what I think was a factor in the draft prices listed: price should be tethered to real performance- the expected SWP/game/million; which I think someone said was about 30 for the purposes of the draft. Price-tethering would be based on the player's expected points/game/price, and the player's cost would increase/decrease proportionally (exponentially?) based on how far from that expected number the player's actual point/game/price was. So, the player would have an expected "base SWP/game/mil" based 50% on the last four games and 50% on the previous 16 games (approx. 1 season), and a "real SWP/game/mil" based on their actual last, say, five games, in terms of their points/game/mil, their price would "tether" back to some constant. Perhaps the constant should be tied to a postitional element (like defenses should average 200 pts/game/mil, QB's 300). But like I said, if their 'real performance' wasn't too far from their 'expected performance,' then the price change should not be too drastic. Hight percentage of ownership of a player should always drive price up- a lemming/human intuition factor. I guess what bothers me most is that Favre is the third most expensive SW player, Fred Taylor is the 8th most expensive player, and Mike Anderson is like the 300th most expensive player. A player like Anderson who comes out of no-where to be AWESOME should skyrocket in value, and a player like Fred Taylor or Favre should sink like a rock, until they begin to put up appropriate numbers. Anyway, this idea may have been put forth in similar or in very different terminlology, I don't know, but the pricing is one of the most interesting aspects of the game, so I thought I'd toss my ideas out there. | |||||
| 14 | Guru ID: 330592710 Fri, Sep 29, 09:37 |
Price tethering. A clever term. There are some conceptual problems, however. First, SW has always provided a market price mechaism tied solely to market forces. I doubt if they would even be interested in departing from this tenet. Second, a tethering scheme has the disadvantage of looking backwards. Mike Anderson is a good example. Looking backwards, you could support a significantly higher price, probably among the elite at the running back position. But we also know that if/when Terrell Davis returns to full strength, Anderson's stats will probably decline. For that reason, he should continue to be priced at a discount. For every player whose price might be "improved" via a tethering approach, there is probably another player who's price would be distorted. I once suggested an approach whereby gravity and anti-gravity determinations might include a performance test. I'm not even sure that is a workable approach, however. Again, it might create as many anomalies as it solves. But I think that's as far as I would ever recommend going with performance based measures. If there is any material trading flow in either direction, performance measures shouldn't be a factor. The current approach tends to produce several players each season that are chronically underpriced. These are the sleepers who come from nowhere to produce big numbers. Two years ago these included Randall Cunningham, Doug Flutie, & Randy Moss. Last year they included Kurt Warner & Stephen Davis. Most competitive teams own them. And anything done to push up their prices adds to the general inflation. But since they are so widely owned, it doesn't really make much difference in the relative values of those teams. So even correcting the most obvious mispricings would be likely to have some unwanted consequences. | |||||
| 15 | Beta Alpha Sigma ID: 33638268 Fri, Sep 29, 11:58 |
i think the best thing SW did was cap the gains by a mil either way-up or down. I remember Randall two years ago jumping 3 million in a week. That was nuts. I made the money but it just made it too easy to make $. This year is the hardest yet. Four price updates and still most teams are under 60 mil. IT makes the money factor more important than it has been in a while because if you only focus on points-youll get left behind the second part of the year. | |||||
| 16 | StLCards ID: 238421814 Fri, Sep 29, 12:40 | One thing I've always wondered is the effect of non-valid rosters on price movement. While I have not done a statistical sampling my recollection is that there is a large number of managers that start drafting teams, run out of money, and never finish. These teams usually have the big names on them. If these invalid rosters all had Player X for instance is Player X then considered widely owned and thus minimize his gravity effect even though he is performing poorly and few active players own him? | |||||
| 17 | mad scientist ID: 528262913 Fri, Sep 29, 13:27 |
I wouldn't recommend any system which was only price-tethered. Maybe the 'tether factor' should only kick in for people who are dramatically over or underperforming their expected pts/game/price, and should only draw them back to within a range, say 15-50 swp/game/million. After they are within the range, then 'regular' (gravity, trades, anti-gravity) market forces would be the only factors. If the prices were at least closer to productivity, it would put more emphasis on picking up good matchups and trade preservation to get those matchups. I guess I would just like to gamble on Favre w/o being stuck with his exhorbitant cost. If players were priced more appropriately, there would be less roster homogenization (aka, 'must-have players') through-out the leagues, and more opportunities for creativity in our rosters. I guess a temporary solution would just be to increase gravity. | |||||
| Rate this thread: | If you wish, you may rate this thread on scale of 0-5. 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: (But first, how about checking out this sponsor?) | |||||||