These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

Market Discussions

 
  • Topic is locked indefinitely.
12Next page
 

0.01 Isking and Not - Empirical Evidence
Author
Previous Topic Next Topic
Gaston Miromme
The Scope
Gallente Federation
#1 - 2014-09-09 09:23:48 UTC
I often hear station traders arguing against 0.01 isking in favor of more significant price adjustments. These aggressive adjustments are intended to increase trader profit via reduced competition and/or increased transaction volume. The evidence I hear in support of these effects generally takes the form of anecdote or handwavey argument from principles.

Quote:
Simple I have crashed prices down low enough in some markets all my competition picked up and left ... This gave me a good week of having almost no competition and than slow competition compared to before for an additional 2weeks after that



Quote:
The main problem of station trading is the 0.01 curse. ... The worst thing you can do is following the same strategy. ... You shall cut the price agressively. ... The reason to do so is increasing volume. If the sell price is high, many buyers will say "screw you" and go farm it themselves or set up yet another 0.01 buy order. If the buy price is low, the farmers will say "screw you". If the two prices are reasonably close, they accept it and simply sell/buy so many items that all 0.01 orders before you are filled up and yours too. 0.01-ers usually leave at some point because they have to spend lot of time to update orders.


Quote:
No we slash the price until it falls into a range where more suppliers are willing to sell and more consumers are willing to buy. Or in other words we bring it down from prices where the product displays inelasticity to a point where the product starts to behave more elastic.


Does anyone have actual empirical evidence for these effects? Because my initial explorations of the data (CREST history data for a smattering of different modules) does not seem to support these claims:

Graphs of Several Modules

Visual inspection doesn't suggest any clear trends in transaction volume or orders in response to varying spreads (contrary to the hypothesis established by advocates of aggressive adjustments). In each graph, no trend line could be found with an R^2 exceeding 0.05.
Cista2
EVE Museum
#2 - 2014-09-09 11:05:11 UTC  |  Edited by: Cista2
The data do not tell if volume goes towards buy orders or sell orders :) only the daily median give some indications of that.

If I am moving my buy order aggressively upwards, I may shift some of the volume from sell orders to (my) buy orders. (And vice versa, traders most often only have one of the types of orders up).

So, even though in your data you would not see a change an increase in overall volume with smaller spread, I have provoked a larger volume going through buy orders, and therefore got my buy order filled faster.

My channel: "Signatures" -
Elizabeth Norn
Nornir Research
Nornir Empire
#3 - 2014-09-09 17:15:18 UTC  |  Edited by: Elizabeth Norn
Would you be able to tell if someone pushed other traders out of the market using your graphs?

I imagine the 'price goes down, volume goes up' hypothesis works for some items like newly introduced items such as faction ships, exploration modules and really expensive items like deadspace gear, because their prices can be too high for a lot of people.

Free 3rd party services

Buddy link w/ bonus 250k SP, you keep the whole PLEX upon subscribing
Vaerah Vahrokha
Vahrokh Consulting
#4 - 2014-09-09 19:51:02 UTC
Actually, the typical downtrend begins with small volumes.

Volumes only increase after a while, when people start mass panic selling.

What some guys believe to do is smaller than what happens for real and lasts a limited time as well. I am not sure it's possible to catch those spikes with those charts.

EvE Online portal | EvE markets tutorials
Gaston Miromme
The Scope
Gallente Federation
#5 - 2014-09-09 22:36:18 UTC
Quote:
only the daily median give some indications of that. If I am moving my buy order aggressively upwards, I may shift some of the volume from sell orders to (my) buy orders.


Here's a graph I came up with to test your claims. If your claims are true we should expect a large decrease in the high price (compared to the average high over the preceding 7 days) to shift the median price toward the high price and a large increase in the low price to shift the median price toward the low price. In both cases, this means a trend line with a negative slope. If we force a fit, we do get lines with a negative slope. But the correlation coefficients are pretty abysmal.

Quote:
Would you be able to tell if someone pushed other traders out of the market using your graphs?


The orange diamonds aligned to the right axis (total outstanding orders on the market) should give some indication of that. They do not seem to correlate with spread. I suppose one could argue that aggressive price adjustments "push others out" by discouraging them from updating orders which effectively removes from the competition without being reflected in the total number of orders. However, it seems unlikely that, simultaneously, this effect is large and the effect on the total number of orders is essentially non-existent.

Quote:
I am not sure it's possible to catch those spikes with those charts.


As in you believe they are intra-day events? It seems that anti-0.01 ISKers also advocate infrequent price checks (~1/day). Given this, it seems they're not well-positioned to take advantage of intra-day events.
Cista2
EVE Museum
#6 - 2014-09-10 09:55:51 UTC
If it is discussion you want, you desperately need to explain the bases of your tests better [Shocked ... we are not inside your head you know Big smile

My channel: "Signatures" -
Vaerah Vahrokha
Vahrokh Consulting
#7 - 2014-09-10 13:02:48 UTC
Gaston Miromme wrote:

As in you believe they are intra-day events? It seems that anti-0.01 ISKers also advocate infrequent price checks (~1/day). Given this, it seems they're not well-positioned to take advantage of intra-day events.


The market PvP your study describes, often involves "action and reaction" patterns. A guy 0.01 ISKs for 1 hour, the other guy reacts dropping price by 100k. That's 1 hour, 2, maybe 4, so yes, it'd say it's mostly an intraday activity.

EvE Online portal | EvE markets tutorials
Bad Bobby
Bring Me Sunshine
In Tea We Trust
#8 - 2014-09-10 13:20:52 UTC
I think that the effectiveness of psychological warfare in trading has less to do with the market spread and more to do with the aggregate psychological vulnerability of your competitors.
Gaston Miromme
The Scope
Gallente Federation
#9 - 2014-09-10 20:46:02 UTC
Cista2 wrote:
If it is discussion you want, you desperately need to explain the bases of your tests better [Shocked ... we are not inside your head you know Big smile


Okay, I suspected as much. The general idea is that advocates of aggressive adjustment are making causal claims:


  • If you reduce the spread aggressively (cause), the transaction volume will increase (effect)
  • If you reduce the spread aggressively (cause), you will push out competitors (effect)


Since causation implies correlation, if these claims are true, there should be a correlation between the relevant variables (e.g. spread and transaction volume) in the historical data. Absence of correlation (or weak correlation) provides evidence against the truth of these claims.

In actual fact, example graphs do show only minimal correlation. This suggests that the original claims are false.

Does that make things clearer? Or are your uncertainties about particular graphs? In my opinion, the follow-up graph I made in response to you (regarding the median price) is the definitely the most complicated.
Cista2
EVE Museum
#10 - 2014-09-10 21:24:56 UTC
Gaston Miromme wrote:
In my opinion, the follow-up graph I made in response to you (regarding the median price) is the definitely the most complicated.

Yes that's the one i couldn't be bothered to look at without any guidance.

My channel: "Signatures" -
Gaston Miromme
The Scope
Gallente Federation
#11 - 2014-09-10 22:08:22 UTC
You saw the little text description under the graph, right?

The idea is that the x axis reflects changes in high bids (low asks). If the high bid had been at a constant 100 ISK for the preceding 7 days, until someone enters a bid of 120 SK today, that would be a 20% increase or 0.2 on the x axis.

Then the y axis is a proxy for the type of transactions completed (Are all transactions completed by people buying from sell orders, selling to buy orders, or some mix of the two?). Because this information is not directly available in the historical data, we instead use the median price as a proxy. If the median price is very near the lowest recorded transaction price for the day (the bid price), most transactions took the form of selling to buy orders. If the median price is very near the highest recorded transaction price for the day (the ask price), most transactions took the form of buying from sell orders. By finding where the median price falls in the range from low to high price, we get an indicator of the relative proportion of buy fulfillments and sell fulfillments.

These two things together allow us to test what I take your hypothesis to be: Aggressive increases in buy order prices may increase a trader's volume (while maintaining overall transaction volume) by encouraging the fulfillment of buy orders instead of the establishment of sell orders. If this hypothesis is true, we should expect (for each of the two series---one showing response to buy price increases and one showing response to sell price decreases) a strong negative correlation. As the bid price increases, the fulfillment of buy orders will be a higher portion of transactions (pushing our proxy, the median price, down toward the high bid price).

Is that any clearer?
Pookoko
Sigma Sagittarii Inc.
#12 - 2014-09-18 02:14:39 UTC
+1 for the effort to attempt methodical analysis of the 'claim'. But as Bad Bobby pointed out above, I'm not sure if this 'psychological warfare' is something that can be measured in quantitative sense.

This method cannot account for the case where particular competing traders 'pull out' of a particular market segment. Also, the volume of trade does not show how many traders are involved in the market, only how many transactions are made.

Number of traders involved in the market can change, but transaction volume can remain the same. Also, not all market activities are attributable to trader actions. Much of the volume traded can be attributed for 'consumers', whether they buy through placing their own buy order or buying off sell order.

We have no way of deducing this using any statistics available through in-game data. In conclusion, because the primary goal of 'aggressive' order adjustment is to fight other 'traders', without being able to figure out what transactions are from 'traders' and what transactions are from 'consumers', it is impossible to draw conclusion from quantitative data whether 'psychological warfare on other traders' have impact.

Where this fails, we need to go by the qualitative data based on the experience of actual traders who trade aggressively, or who leave a market segment because of actions of other traders. The question is not 'how many', which you may be able to plot in a graph, but it is more the case of 'does this actually work on some people?'. And for this, the answer, from my own experience and experiences described by other traders in this forum, seems to be 'yes sometimes'. Again, we cannot quantity on how much percent of traders this works/doesn't work, but we can at least make a claim that 'it works for some'.

In this case, what you may aim to test as an 'assertion/claim' using quantitative method, can be turned around and regarded as an 'empirical data' in itself - account of experiences told by traders.
Ocih
Space Mermaids
#13 - 2014-09-18 19:49:09 UTC
I've had 2 of my trades get pushed way above the .01 ISK war and all I see are bubbles.

The demand still isn't there for what I trade in.

.01 ISK is a market game and is born of necessity.
Tinu Moorhsum
Random Events
#14 - 2014-09-19 14:28:28 UTC
Here's a question (proof it left to the student).

What would happen if you simply didn't play the 0.01 isk game?

T-
Rockerguy 101
Deep Core Mining Inc.
Caldari State
#15 - 2014-09-27 19:41:39 UTC  |  Edited by: Rockerguy 101
I just started trading with my alt about a week ago and I really can't understand the desire to .01 isk orders. The way I look at it a trader's purpose in life is to help merge the buy/sell orders together so other players can obtain the items at the fairest price possible. Making isk during the process is the reward for doing so. If you manage to reduce the trading profit to the point where it's no longer profitable enough to trade then just move on to another item or maybe even temporarily change your trading hub if you are in a secondary market.

Explorer  |  Ghost | Caldari Militia
Jace Sarice
#16 - 2014-09-28 04:30:32 UTC
Rockerguy 101 wrote:
The way I look at it a trader's purpose in life is to help merge the buy/sell orders together so other players can obtain the items at the fairest price possible. Making isk during the process is the reward for doing so.


Nonsense. A trader's purpose is to profit. End of story. There is no altruistic hippy motivation behind ISK.
Rockerguy 101
Deep Core Mining Inc.
Caldari State
#17 - 2014-09-28 14:38:15 UTC
Jace Sarice wrote:

Nonsense. A trader's purpose is to profit. End of story. There is no altruistic hippy motivation behind ISK.


As a matter of opinion I can't say I entirely agree with you, but I think you are correct about the motivations of traders. All I know is as a new trader I'm sure as hell not wasting my time .01 isking orders. So far it's worked out just fine.

Explorer  |  Ghost | Caldari Militia
Jace Sarice
#18 - 2014-09-28 14:50:01 UTC
Rockerguy 101 wrote:
Jace Sarice wrote:

Nonsense. A trader's purpose is to profit. End of story. There is no altruistic hippy motivation behind ISK.


As a matter of opinion I can't say I entirely agree with you, but I think you are correct about the motivations of traders. All I know is as a new trader I'm sure as hell not wasting my time .01 isking orders. So far it's worked out just fine.


Strategies vary, sure. I wasn't commenting on the pennyfracking topic - just the notion that traders are driven by some sort of love for fair market prices.
Pietro Vitale
Better Off Red
Unspoken Alliance.
#19 - 2014-10-15 20:30:42 UTC  |  Edited by: Pietro Vitale
Gaston Miromme wrote:
Here's a graph I came up with to test your claims. If your claims are true we should expect a large decrease in the high price (compared to the average high over the preceding 7 days) to shift the median price toward the high price and a large increase in the low price to shift the median price toward the low price. In both cases, this means a trend line with a negative slope. If we force a fit, we do get lines with a negative slope. But the correlation coefficients are pretty abysmal.

Fascinating study.

I'd be interested to see the y-axis modified to show the average change in ratio over a period of time (2 days, 5 days, 7 days etc.) after the price change, rather than just the ratio. Right now you are aggregating the cause, but not the effect, basically assuming that responses to your manipulation happen semi-instantaneously (same day). Averaging the ratio over a time period should also filter out some of the "random noise" from consumers that don't care about the margin.

That is assuming I'm correctly understanding your plot. Big smile
Ioci
Bad Girl Posse
#20 - 2014-10-17 05:17:48 UTC
A few years ago I had a Providence Freighter in Essence region. I wasn't certain I wanted to sell it but it was at the time when they had spiked over double what I paid for mine for some reason so I watched markets. I checked Essence market daily for 6 weeks and no Providence Freighters were on the market. The usual 5 mill buy order garbage but none for sale. Every day for 6 weeks. I put mine up for sale, in less than 15 minutes one was on the market for a mill less. I didn't bite, I pulled the order, flew to Jita and sold it on the buy order. It was win, win for me. I was going to make twice what I paid in any case but in the time it took me to fly it to Jita and return, the block order had also been taken down.

That is a typical situation in every order in every region of EVE. It's called market PvP for a reason. They are trying to start price wars, down or up to get you caught up in an irrational competition where they bluff you up or down so they can pull your stock at a profit.

.01 ISK'ing is a taunt. Nothing less, nothing more.

R.I.P. Vile Rat
12Next page