| From | Message |
|---|
|
 My bitch of the year (2020)
Team match declined: veronfr1 vs biblio5 - without looking at team names, who has the
advantage. Please elaborate.
|
|
To me, that looks quite even, I'd take that match. That's me though, I basically look at the average 90 day ratings, then the current and top 90 day. If that match would be declined, I would assume it's due to a parameter I'm not so familiar with. After looking briefly, it'd be hard to decline, but again that's just me.
|
|
After some conversations, that pairing was ultimately accepted. It was a mistake declining initially.
|
|
Ah okay good, that match is very good, I would have been surprised if it wasn't an accidental decline.
|
|
Looks like a very fair match
|
|
I would give the advantage to veronfr1 based on head-to-head records (7-4-1). But I would not decline it.
~Tim
|
|
... continuation of bitch of the year.
I think what caused the "miscommunication" was that I requested one of two players for biblio5.
A new one: qr19kaash vs michael1999, who has the edge and by how much?
|
|
qr19kaash, based on his average opponent rating.
|
|
Okay Tim. Very nice "find" on both possible pairings. However, in the first challenge the Captain of
veronfr1 initially declined. Our player in the 2nd pairing got very upset with me for avoiding a
player that has a 13-1 record against him.
I know one Captain that will decline for the reason you mentioned in this pairing. Maybe you are
number two.
Are there any other advantages for either?
|
|
No, I would not decline just based on the average opponent rating. All that tells me is the path that the player took to achieve their ratings. If it is high, then they are challenging themselves. If low, the reverse is true. If it is even to the other stats, they are playing right at their strength.
As to your last question, I don't see any other significant advantages in either matchup.
~Tim
|
|
The reason the 2nd challenge was declined will be stated shortly, after a couple more replies.
I have never heard of this reason for declining but there is a similar one.
|
|
qr19kaash is stronger - the midpoint between his opponents' line and his on his graph is about 1746, whereas michael1999's midpoint is only about 1692. Michael1999's current rating is inflated because he is playing weaker opponents.
|
|
@ Dave
But I thought that was equalized because michael1999 is higher rated on 90 day average, past
year average and he appears still be rising in rating. Also there is a 20 point difference in current
rating, michael1999 higher.
Your analysis is very close to the decline reason.
|
|
@Mike
His rating is rising but his opponent rating on the graph is dropping steeply
|
|
But he very recently won three matches 2-0 against higher rated players pricop,
monkeyinspector and wierdaj His last two wins were against a 1501.
Having said that, your point is valid (qr19kaash is stronger). But by how much?
|
|
Based on my analysis above, 1746-1692,or 54 points.
|
|
Okay Dave, valid points but not sure I fully understand.
A few more things to consider: 1) have you ever had a decline based on highest rating ever on
Gameknot, or 2) high timeout %, or 3) team game rating, or 4) team game winning %?
These may not all apply here.
|
|
Also, most of us know Shirley shirlmygirl and her obsession with FULL rating graph. That is
another very valuable tool for getting a clearer picture of a player's game history.
|
|
The calculation I did in my post at 12:18 today yields a rating adjusted for the current average of opponents' ratings.
I've had a challenge declined because of high timeout % but none of the others you listed.
|
|
That challenge declined due to high timeout % may have been me.
I have had a couple challenges declined based on highest rating ever on Gameknot. Just like
90 day high rating the highest rating are way overrated, IMO. Sometimes, the highest rated player
defeated was by timeout!
Also, and very important, players that have many games and play very fast (as qr19kaash)
will have striking high points (90 day high rating) as well as lows sometimes approaching rating
floor. (see FULL rating graphs).
Somewhere in this post is my hint on why this 2nd challenge was declined.
|
|
mikemate makes a good and overlooked point about high volume players. Their one-year graphs and their 90-day stats contain many more games than the graphs and 90-day stats for slow players.
What would the (shorter) rating graph be for a player who had completed zero games during the previous year? A horizontal line. The easiest way to show rating consistency is to stop playing!
In general I would prefer to be given statistics based on a certain number of previous games (for instance, an average rating for the previous 100 games) than statistics based on months of play.
Most of us aren't going to calculate the "100-game average" ourselves (and I don't either!), but I think we should keep in mind that the 90-day stats for two different players can reflect a highly unequal number of games. And the player with more games will tend to have a wider rating range.
|
|
There is one stat that actually uses the "past 100". It is the average time/move. A very slow
player can be at 2 hrs/move very quickly by playing two or three games very fast.
|
|
Here was the decline message: "Thanks for the challenge. Whilst your suggestion seemed fine
there was actually a 100 point advantage for your player on the rating floor."
First time I saw that message in 11 years on Gameknot.
|
|
I know which team that is. I haven't had that message from them for a while though, and I haven't been looking at that stat. That tells me that only one of the captains there looks at that stat.
|
|
That stat is very similar to highest rating ever which at least one team uses to decline a challenge.
Both stats are nearly irrelevant, IMO.
|
|
I know which team that is as well. I can see their reasoning on why they use that stat, but yeah, for me, I don't see it as a good stat to use (in both cases). Oh well, that's my opinion, I'll adjust to their preferences best I can, but I won't be using it myself elsewhere.
|
|
Mikemate
These days, Mike, I don't look at the FULL rating graph most of the time, but the most current graph. In fact, when I am considering or making a pairing, the very first thing I look at is the current graph on the proposed opponent's profile page. I occasionally look at the extended graph, especially if I think that a particular player is cheating. I think it is probably well known that I go after cheats, usually reporting them directly to GameKnot.
You taught me how to use the rating graph, Mike. I use it often and consistently, and am really grateful that I learned the significance of it from you. For those of you who don't know this, I once was on Mike's team.
|
|
I am too lazy to look at the rating graph all the time. Current, top 90 day, and average 90 day ratings is what I look at first. I also pay attention to time per move (trying to not pair my fast players with slow players) and making sure my players aren’t already playing the player I’m considering, or within the past month (I aim for longer with my highest rated players…I started doing this after a player on mikemate’s team was unhappy with playing my player again after just finishing games. I think it is good to diversify things). Then of course I try to match up preferences the other team has too.
If it’s with a team I closely trust, I’ll try adding looking at ‘Rivals’ history, the average opponent 90 day rating, and if I remember, the rating graph.
Yes…it takes me some time to start matches. Good thing I have a co-captain helping a lot now, it was hard for me to keep up.
|
