Make your own free website on Tripod.com

[Guru Fantasy Sports logo]

Feedback

This page is devoted to feedback I've received from you. If you have a question to ask or a comment to make, send me an email and if I think it's of general interest, I'll post it here.

Click here for more recent feedback items.


Guru:
TO TRADE OR NOT TO TRADE: THAT IS THE QUESTION

As the NBA season runs down its fast-break path, fantasy owners usually have one of two thoughts: "Please don't get hurt" or "Please get hurt". Although this may be pre-conceived, it basically revolves around one important factor for Smallworld owners: To trade or not to trade.

As anyone may gaze upon the Worldwide Leader board, there is a binding relevance to each team, besides that of similar players. Literally 90% of the teams in the top 50 have less than 10 trades remaining, with less than 50% of the games being played. What does this mean? Well, as all of us know, there is no guarantee that any one player will not get hurt. For the most part, the major, major stars have undergone a rash of injuries this year, making for a lot of managers to devise a way to get semi-stars cheaply and grow off their "profit". Then, when they have made significant money, dumping them off when the major star comes back to play. Case in point: many owners jumped on Kevin Willis when Hakeem Olajuwon went down, sending Willis above 7 million, no doubt due to his affordable price and high production. Right now Hakeem is hovering around 5.5 million, and as anyone can guess the Dream will shoot through the roof when he comes back in a few weeks, lowering Willis's status considerably. This is a given.

However, the only people at this point who can afford such luxuries as this wheeling-and-dealing are those who have a significant number of trades left. Because of the 12 man team and the scare of injuries, most will not gamble on Olajuwon eating up one of the precious trades that could come back to haunt them as the season goes on.

So, what's the deal (no pun intended)? Is it wise to trade so much so early? Well, it goes for naught because no one can tell the future in regards to injuries. Typically the best option to make is to have an AC Green on your team rather than Laphonso Ellis just because of the fact that we all know who plays and who doesn't because of injuries. When you're talking the immeasurable price of trades, giving up points in this case for rock-solid confidence of being injury-free can make all the difference in the world. Trading early has helped all of these teams get to where they are in the Worldwide league. Those teams that have an ample number of trades will be able to get Penny Hardaway, Brian Grant, Olajuwon, etc., for bargain-basement prices. How they manage their team after the prices sky-rocket is another issue.

To Trade or not to Trade? Well, this question can only be answered at the end of the season, when the injury gods have to put to rest any stress caused by this oh-so-important facet of the game. Most will say trade away; those are the types that you will hear screaming at the same time when Robinson goes down in pain, or Jordan limps off the court. Then you will see me giggling, as I know I will be able to get both of them dirt-cheap because I've budgeted. My team is not in the top 50, (not even in the top 2,000) but I believe the method of budgeting your trades is the best way to go.

Trading early: it's a gamble, enjoy the ride, and keep both hands inside until it comes to a complete stop.

T. Gossett (1/26)

response from Guru:
If you come back from a rank of 2000 to contend, I'll eat some of your crow. Granted, an excess of 2nd half injuries may torpedo some of the current leaders - but probably not all of them. I think you've got to trade aggressively just to be in contention for the end game. From then on, a little luck wouldn't hurt. - Guru


Guru:
Why don't you add standard deviation to the summary stats page?- D. Milet (1/24)

response from Guru:
I don't think standard deviation is important enough to include on the summary stats. If you just look at a player's individual game-by-game stats, you can see whether he's someone who is consistent or not. Using standard deviation was just a useful tool to identify those players who were at the two extremes. - Guru


Guru:
The two biggest first-half surprises have to be the two rookie point guards who have featured prominently on successful Smallworld teams: Brevin Knight and Bobby Jackson. Other point guards, notably Chauncey Billups and Antonio Daniels, were drafted higher, but Knight and Jackson have blown these two away. You could argue that Jackson's success just comes from getting playing time on the worst NBA team ever assembled, but Knight is the real deal. He's made the Cavaliers forget about Terrell Brandon already. He even leads the league in steals. Not to mention all the profit he generated for active smallworld managers with his early season price fluctuations. - N. Mason (1/23)

response from Guru:
Knight's price gyrations early in the year were instrumental in allowing my team to amass a fortune. It was like taking candy from a baby. - Guru


Guru:
The average "trades remaining" dropped from 4.94 last week to 3.84 this week for the Smallworld top 50 rosters... 30 teams with less than 5 trades left, as opposed to 26 last week. Barkley was on 27 rosters as of 1/5, so he could account for about 0.5 of the 1.1 change from last week to this week alone (assuming that about 25 out of the 27 managers sold him... I haven't checked it at all though).


 Trades      # of Rosters   # of Rosters           
Remaining      THIS WEEK     LAST WEEK      Delta
   0-1             10            7           + 3
   2-4             20           19           + 1
   5-7             17           11           + 6
   8-11            2            12          - 10
   12+             1             1             0

- D. Taylor (1/12)

response from Guru:
Interesting that 10 out of the top 50 have no more than one trade left. Looks like injuries could play a major role in determining the ultimate winner.- Guru


Guru:
I came into the SW season having not fantasy hooped since Kiki Vandeweghe was a good pick, and had only a few days to prepare after hearing about SWHoops... So, I took a quick peek at the scoring system and decided that trying to win through scoring lots of (realworld) points wasn't the answer, because of the missed FT and FG subtractions. So I scouted my draft through adding only players that would have an RASB per game of ten or greater (rebounds+assists+steals+blocks). Despite the fact that this "formula" uses only half the variables, and none of those in the right proportions, it has worked well enough to put me in the top 300. Notice that Grant Hill, Shaq, the Admiral, Chris Webber, Gugliotta, Garnett, etc have both the big SWP and RASB totals. And guys like Marbury, Brevin Knight, Camby, Brian Grant, etc have that threshold RASB of around 10 per game, and they are also about the least valuable SW players you can have and still succeed- AirSteve3 (1/5)

response from Guru:
An interesting shortcut. Of course, now that you have fully developed stats available here, I suppose there's less reason to need the "short form".- Guru


Guru:
This is a follow up comment on your Hoop Pointers essay The Price is Left and our subsequent dialogue in th feedback section.

OK, I think we both agree that there needs to be some "subjectivity" in deciding who to select for our teams. But for the purposes here, I am just dealing with the objective side of things. Another way to look at this is to say I am considering the graph of a Player's Price vs. My EXPECTED SWPPG for that player, which allows me to take others factors into account (such as # of games to be played, chance of injury, etc.).

OK, so the problem I have with a curved line is this. For the moment, I'll assume a simplified roster: 2 players, $11,000,000 salary cap. According to the green line in your second hoop pointers essay, I should be willing to pay about $500,000 for 10 SWPPG (rough estimate). I would also be willing to pay about $1,500,000 for 15 SWPPG. At $9 million, I expect 38 SWPPG. Finally, at $10 million, I expect roughly 40 SWPPG. OK, so if this is the efficient frontier, and all 4 players we are considering lie ON the frontier then I should be indifferent between these players, because they are all most efficient, and to maximize SWPPG I should therefore simply attempt to spend as much of my $11,000,000 as possible. But if I buy the $9 million and $1.5 million players, I end up with 3 SWPPG more than I would with the $10 million player and the $0.5 million player. The example can extrapolated even further, in that two $5 million players at 28 SWPPG bring in another 3 SWPPG on top of that.

So, what I am saying is that while I think the green line is "efficient" in terms of describing the actual prices of players, this does not make it efficient for the purposes of maximizing your SWPPG. Perhaps this is the difference between our approaches. In terms of maximizing SWPPG, only a straight line makes sense as an indifference curve. But if you are looking at pricing and attempting to maximize your salary cap, then your frontier makes sense. The difference is subtle but significant. I think this is why I was confused before, and I am just now realizing it. You were looking more at the left-to-right deltas, along the price axis, attempting to maximizing salary cap growth, whereas I'm looking at up-and-down deltas, along the SWPPG axis, attempting to maximize SWPPG. Ah-hah. Well, now that is cleared up, and I probably don't need to send you this letter now anyway, but oh well. I typed it, might as well send it.

Oh, just one more observation on the frontier though. I think many of the players in the sub-$2,000,000 price range are there simply because nobody owns them, so they can't go any lower. But I don't expect them to rise in value (move right). But they therefore skew that portion of the line, and I would not expect a 12 SWPPG player at $200,000 to necessarily rise in value, whereas at higher values (a 25 SWPPG player at $2 million, for example) I would expect their value to more towards the frontier.

OK, enough of all this. - D. Taylor (12/23)

response from Guru:
You've probably thought more about this than I have... or anyone else, for that matter. You're right when you say that my focus was on identifying undervalued players (which is why I titled it "The Price is Left"). You are also right when you say that many players will never appreciate to their "fair price"... often for reasons highlighted in my subsequent Price Fixing essay. And you are right when you say that using a curved frontier does not necessarily maximize SWP/G. (In fact, there are usually a few lower priced players who are far enough "to the left" that the ultimate efficient frontier really does look more straight than curved.) And most importantly, you are right when you say "enough of all this." - Guru


Guru:
I created a program that calculates the number of Small World Points when the user types in the stats. Anyone can download it at http://members.aol.com/tarzan99/ptsadder.zip (it's a zip file). - Josh Dunkleman

response from Guru:
Nicely done, Josh. I'm sure many will take advantage of your program. - Guru


Guru:
New free basketball league starting up!
Title says it all, I don't know if you dislike the cnn-si format or the league itself. But, they do have a midseason league starting up and everyone needs to have their draft lists set up by January 14. The site is http://www.sandbox.net/basketball/pub-bin/frame?ss_nba_main+main.html . - C. Mabry (12/31)

response from Guru:
I haven't checked into the CNN-SI leagues yet, although I've noticed that a number of you who have registered have indicated that you play that game. I've been so focused on getting my site up and humming that I just haven't had the time to pursue. Even my ESPN team has been backburnered while I organize the second chance activity. But thanks for the alert... I'm sure others will take advantage. - Guru


Guru:
I would like to see the top ten or twenty highest single game Smallworld points-both with and without triple doubles (only three at last count). - L. Toth (12/23)

response from Guru:
Interesting idea. While I always review the top games each day, I hadn't been keeping track of the top single games by SWP for the season. So I did some sifting, and here is the list of the top 20 to date. As you can see, it now requires 61.5 to make the top 20, including ties. I also found it curious that two of these games occured on opening day (10/31 - Kidd and Garnett). I'll update this table as the season progresses, so I've added a link to it in the left menu. - Guru


Guru: (See also a later related discussion above.)
Personally, when comparing player prices, I prefer to use a straight line to a curved frontier. The reason is that by using a straight line, you can accurately compare the difference between combining an expensive and cheap player as opposed to two medium priced players. Basically, what I do is take the line:

Required SWP/G = 19 + (price in $millions) × 2

That is, for every $1m increase in price, I require a return of 2 additional SWP/G, so that a $2m player needs to score 23 SWP/G, a $6.5m player needs to score 32 SWP/G, etc.

By comparing my expected SWP/G for each player to my required SWP/G, I calculate a delta. If the delta is +2 or higher, then I tend to buy. And likewise I tend to sell for deltas of -2 or lower. I don't worry so much about maximizing total points... I'd rather have a $2m player who I think is worth $5m and leave the excess cash laying around than have a $8m player who I think is priced appropriately, and whose value I don't think will appreciate. The reason is that within a week, if my theory holds true and the value rises, I can then sell and buy an $11m player, who will very quickly make up the lost difference while I held onto the undervalued player and some excess cash.

I think it would be interesting to examine the cash reserves of top teams in the early parts of the season, although by now most teams are starting to trade for the "long haul" and so are less likely to have unused cash laying around.

In any case, these are all guidelines, and must be tempered with the player's health, # of upcoming games, gut instinct, and other factors. But I think it is interesting to hear about different people's systems. I am new to this game, and I didn't figure out the importance of trading until a week's worth of trading had already passed. Thus I missed some key opportunities. But with about 32 trades used I have reached the $100,000,000 mark and am hovering around 200th place. - D. Taylor (12/23)

response from Guru:
Let me comment on some of your ideas -

1) You may prefer to use a straight line, but that doesn't mean it is. Frankly, there is enough price dispersion around any formula model that it's critically important to look at all of the factors - as you correctly point out toward the end. But, as I discuss in several of my Hoop Pointers, I do think there are some fundamental reasons why the graph, in a perfectly efficient market, would not indicate a linear (i.e., straight line) relationship.

2) Your trading strategy is basically sound, especially for the early part of the season, when it's clearly more important to accumulate wealth than points. But, you have to be wary of all of the factors that drive price. Some players which seem undervalued are likely to stay undervalued. So, you need to decide whether you are willing to hold a cheap player if the market doesn't respond as you expect.

3) So far, it's looking like injuries are going to play a major role in this year's Hoops race. If so, that makes trade conservation very critical. And it also creates profit opportunities that otherwise might not occur in mid-to-late season. I've used 37 trades already, so this is the biggest issue I'm pondering today. Hmmmm. - Guru


Guru:
I wonder if you could set up an area for people looking to change divisions. When everyone in your division gives up there is nothing left to play for. I stop giving it my best. I would like to get in a division where people are serious. - J. Nixon (12/22)

response from Guru:
What a great idea! One of the more mundane tasks I've undertaken in the last few weeks is to download all division names, and then to download all team names (so I can email all of you and tell you about this site). It's tedious, but the price is right. In the process, I discovered a lot of leagues which are totally empty. I selected a few which are not password protected, and which have socially acceptable names. I'll list them here, and suggest a range of rankings as a general guideline for each. If you want to join a division with others who have a similar ranking, feel free to take advantage. I'll set up links here so you can see whether anyone has joined yet.

Division: BULLISH ....... for teams ranked 300 or better
Division: Got Skills? ...... for teams ranked 300-600
Division: Brass .............. for teams ranked 600-1000
Division: Great Scott ...... for teams ranked above 1000

If these seem to get full, I'll dig out a few more names. First come, first serve. Just another quality service from.... Guru


Guru:
Any thoughts on why Oliver Miller's price is going nowhere? He seems to be pulling in some impressive stats compared to others at his position. - Dobie 16 (12/22)

response from Guru:
I'll confess I don't know a thing about Oliver Miller, but I looked up his stats, and you're right. He does seem to be posting pretty consistent numbers, and does trade at a big discount to others with similar totals. I have several ideas as to why this may be:

1) He was injured for awhile, so his cumulative SWP total looks low when compared with others who have similar SWP/G averages. If everyone had access to stats like those I provide, this issue would probably go away. But, so far, I'm only getting several hundred hits per day at this site, so those of you "in the know" are still in the minority.

2) He plays for the Raptors, which means he gets little visibility - and probably little respect.

3) My teenage son says "He's fat." Probably not the reason, but then, there are a lot of teenagers playing this game, so you shouldn't totally dismiss their viewpoint, I guess.

I suppose there is also always the chance that most of the world considers him to be an overachiever, and expects his performance to drop. But I'm betting that the biggest reason is my first one. If that remains so, you may have to content yourself with having a cheap producer, but not a source of capital gains. - Guru


Guru:
One suggestion for you: the fact that someone is injured should figure into the averages somehow. Maybe you should show averages two ways - that is, one average SWP/G which includes games missed due to injury when dividing, as well as the way you have it.

The player that is typical of my concerns is Brian Grant. Grant puts up good numbers when he plays, but, that is not that often these days. A groin injury has him missing two of three games last week. That should somehow be apparent when evaluating a trade. Either way I have a new place to go before making any trades. Keep up the good work. - warkol (12/19)

response from Guru:
You're correct that I only count games played. I figure if a player gets injured, you'll probably dump him. So instead, I'm trying to measure the expected output for a player assuming he's active.

You may have a point, though, for the players who typically miss a game here and there, either for a mild injury or just a simple DNP (such as Iverson's recent one game suspension). In general, I think a manager should try to assemble a roster of players who are likely to play every game. But players who are prone to an occasional DNP should certainly have that factored in to any trading decision. You can figure out whether a player has missed any games by looking at his individual page of game-by-game stats. A game not played will appear on a player's daily history page as a blank space in the SWP column, whereas a game played with zero SWPs will show up as a zero.

However, never let it be said that the Guru is averse to new ideas! ...Presto-change-o!!!... Now check out Brian Grant. I added the extra stats and averages at the top of each player's page, so you can use whatever average you think is most appropriate. Thanks for the feedback. - Guru


Guru:
To what extent would your selection to a Dream Team depend upon the number of games remaining on your schedule (i.e., if you extrapolate from past points/game to future TOTAL points)? As of today, most teams have played a similar quantity of games, but a team like Detroit has played 4 more games than Houston, for example, so Hill's points per game are less valuable than Barkley's, for example (if you assume neither will get hurt / have the flu). - V. Davis (12/18)

response from Guru:
I suppose you could "normalize" the SWP/G stat by factoring it up or down to compensate for the differences in games remaining. As of today, I figured out that fewest games remaining for any player are 57, while the most are 62. Maybe that doesn't sound all that meaningful at first, but then realize that it represents a difference of almost 10%. If you're looking to add a player as a "buy and hold", then it probably makes sense to pay attention to this. If you're buying a player as a hoped-for gains opportunity (i.e., you plan to flip the player at a gain sooner than later), then games scheduled during your planned holding period are all that really matter.

You'll notice that I have now listed the number of games remaining in the top section of each player's individual stat page (see Sir Charles, for example). Of course, in his case, you should probably also factor in an allowance for the games he'll sit out due to the wear and tear on his aging frame. In any event, the number is there for you to use as you see fit. - Guru


If you have comments to make or questions to raise, send me an email. I'll post submissions which appear to have broad interest on my feedback page.

As I add new features to the site, I'd be happy to send you an email to let you know. Just take a moment to register.

Guru Fantasy Sports is produced by Dave Hall (a.k.a. the Guru), an avid fantasy sports player. He is neither employed by nor compensated by any of the fantasy sports games discussed within this site, and all opinions expressed are solely his own. Questions or comments are welcome, and should be emailed to Guru<davehall@home.com>.