GeekGold Bonus for All Supporters at year's end: 1000!
8,176 Supporters
$15 min for supporter badge & GeekGold bonus
21 Days Left

Support:

Recommend
15 
 Thumb up
 Hide
14 Posts

BoardGameGeek» Forums » Game Groups » New Zealand

Subject: Math trade analysis - Statistics, behaviour and dynamics rss

Your Tags: Add tags
Popular Tags: [View All]
Andrew Rae
New Zealand
Wellington
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
I had a few moments on my hands and a very good reason to want to compare some simple outputs in excel and access. So I have put together a few interesting results stemming from the recent NZ Math trade.

Admittedly the New Zealand community is not vastly experienced in math trades, however the results suggest that experience is not necessary to enjoy such an event. Benchmarking comments from other math trades would be useful additions to this discussion.

Traders
188 items were listed for trade, and of these 105 (56 percent) were traded. I consider myself distinctly average having traded 6 out of ten games. And I am not alone. Of the 39 participants 48 percent listed five or more games, slightly more than half (56 percent) listed four or more games.

Table 1. Owners participating by number of games offered



My brief analysis showed that only one person trading more than two games traded all of them, and so my award for the best trader goes to ASTRASTRIKER, who I met briefly to trade games. He was a very nice guy despite being a Man United supporter and happens to think Tom Brady and Kevin Garnett are among the most blessed people on the planet. A well deserved award, you will see a very small deposit of geek gold coming your way soon. By contrast our most demand trader was ASTA who only traded one of his five offerings. Two people did not trade, but their offerings totalled only three games.

Games trading
As we mentioned 56 percent of games traded, a fine achievement but 44percent did not. There are a range of reasons why games may not trade.
- there may not be demand for the item (not wanted or duplicate items);
- the asking price may be too high;
- there may not be enough games in a similar value bracket; and
- the owner asks a reasonable price but another similar offer increases the number of trades.

Of my failed trades, two items fitted in the first category and two in the second. And I think this is indicative. The overall BGG ranking of games in the math trade was 833. Traded games averaged a ranking of 627 while those not traded had an average ranking of 1236. It doesn’t take a genius to tell you that putting up low ranked games is almost certainly a recipe for disappointment.

So it is not surprising that our aforementioned ASTRASTRIKER had an average ranking for his 6 offerings of 152. Honourable mentions also go to ALEXGRANT who had an average rating for his six offerings of 253 and also JOHNNYG who offered up Agricola and Twilight Struggle. These contributions really made the trade.

Table 2. Games traded by ranking group



As part of the math trade parameters I had suggested that the moderator restrict the games in some way based on rankings. A 1500 rank limit was selected which was later amended to 2000. A quick look at the descriptive statistics suggests that 1500 was not a bad suggestion as the traded percentage drops significantly to approximately 1 in 4 games trading for games with ranks of above 1500.

Had the initial rules been adopted there would have been 29 less games in the trade and 8 would not have traded. I was grateful not to have to troll through many games which I was unlikely to be interested in and unlikely to trade for, but the numbers involved suggest that there may be a demand for a low ranked games math trade. After all, one mans junk is another mans treasure. But I support the initial restriction to 1500 which turned out to be quite an insightful estimate. I was certainly surprised that as many as 8 lowly rated games did trade. I'm married and so I'm used to being wrong!

More importantly every game ranked 100 or less traded when you exclude the two editions published before 1981 (Civilisation and Acquire). I suspect Civilisation didn’t trade because too high a price was set, because another copy did trade. If you want to trade a top 100 game and be reasonable about your demands, then you are certain to come away happy.

To my mind there is not a significant difference between the next four categories, they all traded at similar rates; though on balance there is a definite trend to be seen.

Figure 1. Percent of games traded by ranking group



Data
If you are the type of person that is interested in the data then I have summarised the results in the table below, or you can geekmail me with a real email and I will send you the excel file.

Table 3. Summarised Data


Thoughts?
9 
 Thumb up
2.00
 tip
 Hide
  • [+] Dice rolls
Simon Woodward
New Zealand
Hamilton
flag msg tools
badge
Summertime is Christmas time
Avatar
mbmbmbmbmb
In the first table, are the column headings (1) number of games (2) number of people?
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Andrew Rae
New Zealand
Wellington
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
manukajoe wrote:
In the first table, are the column headings (1) number of games (2) number of people?


Yes *looks shamefully to the floor while scuffing his feet*
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Simon Woodward
New Zealand
Hamilton
flag msg tools
badge
Summertime is Christmas time
Avatar
mbmbmbmbmb
citylife wrote:
manukajoe wrote:
In the first table, are the column headings (1) number of games (2) number of people?


Yes *looks shamefully to the floor while scuffing his feet*


No worries! Nice article! Exclamation mark!
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Brendon Russell
New Zealand
Auckland
flag msg tools
badge
Avatar
mbmbmbmbmb
Love me some stats! Thanks Andrew.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Dean Adam
New Zealand
Auckland
flag msg tools
Avatar
mbmbmbmbmb
Hi Andrew, just wondering if its worth separating out trades that failed, versus trades that were effectively withdrawn i.e. I put Chinatown up, but after having a fantastic time with it, put in a null trade option...

i.e. of the games that weren't traded, how many were a result of the game owner not selecting potential trade items for that game?

Which may address your third point:
- there may not be enough games in a similar value bracket; and

not sure, just seems a slightly different type of trade fail between I decided there wasn't anything I wanted to swap it for, compared to I was happy to trade it for x y and z, but the maths just couldnt make it happen...

1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Andrew Rae
New Zealand
Wellington
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
moonglow wrote:
Hi Andrew, just wondering if its worth separating out trades that failed, versus trades that were effectively withdrawn i.e. I put Chinatown up, but after having a fantastic time with it, put in a null trade option...

i.e. of the games that weren't traded, how many were a result of the game owner not selecting potential trade items for that game?

Which may address your third point:
- there may not be enough games in a similar value bracket; and

not sure, just seems a slightly different type of trade fail between I decided there wasn't anything I wanted to swap it for, compared to I was happy to trade it for x y and z, but the maths just couldnt make it happen...



A good point, but the preference lists weren't public, I only used the results, so there was not a lot more definition I could provide. I would note however that games that were withdrawn were not in the results list, but null lists weren't marked.

To my mind a null trade just means it is valued too highly. The higher you value your own games the less likely you are to trade because the range of games in your bracket reduces. Thats okay if you value it and want to keep it, but it just means your probability of trading was quite low because you valued it so high relative to other games that you may already have or not like the look of.








.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Craig Liken
New Zealand
Christchurch
flag msg tools
Avatar
mbmbmbmbmb
For the record Andrew, I managed to trade all 3 games (well 4 actually as I offered one as a bonus) that I put up.

Also I couldn't figure out what your first table meant at all, and as you know I am well used to looking at tables.

Interesting comment about the lower ranked games ie 1500-2000. Not surprsing that the untraded games tended to rank somewhat higher, but the ones that did trade may well have facilitated more trades so I think the whole thing was pretty well done really.

As Dean said it might be good to know how many games didn't trade due to the null trade option.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Dean Adam
New Zealand
Auckland
flag msg tools
Avatar
mbmbmbmbmb
Not sure about 'too highly'!! Its just that there wasn't anything else as good

To an extent you're right, but I'd argue that there are a number of more subtle effects working together - we're more likely to value something we've already invested in, to rationalise our investment so to speak, but then we usually made that investment for a reason, so our collections are largely skewed to our interests...

so I'd kinda argue that a game from my collection, even a game I'm prepared to trade is often going to suit my interests (and be rated more highly by me) more than a random collection of games that other people don't want to hold on to....

The good trades happen where there's a nice intersection between interests. Guess that old saying comes to mind, one man's garbage is another man's treasure.... but I'd caveat with, sometime another persons garbage is well... garbage...

but then that's sort of what your stats are saying too.... it was the good games that got good trades...

blah blah blah, why is it that work is never as distracting as conversations about games?


citylife wrote:
moonglow wrote:
Hi Andrew, just wondering if its worth separating out trades that failed, versus trades that were effectively withdrawn i.e. I put Chinatown up, but after having a fantastic time with it, put in a null trade option...

i.e. of the games that weren't traded, how many were a result of the game owner not selecting potential trade items for that game?

Which may address your third point:
- there may not be enough games in a similar value bracket; and

not sure, just seems a slightly different type of trade fail between I decided there wasn't anything I wanted to swap it for, compared to I was happy to trade it for x y and z, but the maths just couldnt make it happen...



A good point, but the preference lists weren't public, I only used the results, so there was not a lot more definition I could provide. I would note however that games that were withdrawn were not in the results list, but null lists weren't marked.

To my mind a null trade just means it is valued too highly. The higher you value your own games the less likely you are to trade because the range of games in your bracket reduces. Thats okay if you value it and want to keep it, but it just means your probability of trading was quite low because you valued it so high relative to other games that you may already have or not like the look of.








.
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Louise McCully
New Zealand
Auckland
flag msg tools
badge
Avatar
mbmbmbmbmb
liken@xtra.co.nz wrote:

Interesting comment about the lower ranked games ie 1500-2000. Not surprsing that the untraded games tended to rank somewhat higher, but the ones that did trade may well have facilitated more trades so I think the whole thing was pretty well done really.



This would be an interesting idea to test... though a bit of a pain to prepare the lists minus all of the lower ranked games to run through the program again.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
John Shortall
New Zealand
Christchurch
Canterbury
flag msg tools
Avatar
Thanks Andrew. A very compelling overview. I am of the opinion of nothing ventured nothing gained. I put up 14 games and 9 traded. I think this was a fantastic result as I 'hedged my bets'. BGG trading tends to favour an individual fan base looking for their 'grail' games naturally. I look to a wide variety of games for family use and with all ages. This is where overall BGG ranking doesn't bother me so much. BGG has wisely provided separate ranking systems for family, children, abstract, strategy, thematic and wargames. This lends itself to Math trades of different types I suppose.

We could throw a math trade out there for ranking over 2000 whistle, or with a family or strategy bias (that's just me being anti).

Being a visual person do we get more graphs and pie charts? meeple
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Ian Anderson
New Zealand
Wellington
flag msg tools
Avatar
mbmbmbmbmb
Doesn't your thesis depend on the idea that there is a close correlation between our personal ranking and BGG ranking. I'm sure it is moderately close for most of us but the difference between one person's ratings and another is part of what makes the Math Trade work.

For what it is worth our personal correlation is here.
(It looks like there are quite a few games I can trade to you, Andrew -- if I owned them whistle)
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Steven De Toni
New Zealand
flag msg tools
Avatar
mbmbmbmb
citylife wrote:


Thoughts?


Are you sure your not working for Big Brother aka Face Book, that analysis seem awfully close to marketing :-)
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Tony Ackroyd
United Kingdom
Brighton
E Sussex
flag msg tools
badge
Avatar
mbmbmbmbmb
Great stuff!

I did some analysis of a couple of Maths Trades for my geeklist:A 'Math Trade'? WTF is that?
I have added this thread to that list.

I also looked at what did/didn't trade out of the Most Wanted and "Hardly Wanted" (the games that only one person wanted). Could you put some stats up of that? (You may only have this data if you used OLWLG).
I also looked at whether being on the first page or the last page had any influence on the trading success rate.

56% is a great result by the way. Most UK/Europe Maths trade are in the 30-40% items trading range. It makes me wonder whether you are onto something with this limit on rankings - cuts out a lot of the crap.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.