maybe timeout % should be the timeout ratio in the last 3 months?
Clear timeout history after some time

I'm thinking that we should have two timeout ratios? One for all time and the other for the past 1-3 months so they get a chance to play in tournaments still.
ADK
I like Erik's idea of a time based (3 months or year) time out ration. I would also have something where if someone has a high time out ration you would limit the number of games they could have active at a time. If someone is playing 100 games at once and times out on all of them. he could only have 50 games going at once until his time out ration falls --just a thought.

@erik 3 months is some option. Another is counting "happenings", not games. Yet another using some weighted formula which detoriates in time.
Probably the best solution would be to rethink the whole process, drop those percentage based restrictions completely and apply penalties like those mentioned by @aansel instead ("as you timed out a few games recently you are now restricted to 1 tournament, the restriction will be lifted once you finish 10 games properly" - or sth like that)
@aansel Note that if you limit count of games played and keep percentage ratio, you make improving the ratio even more difficult.

maybe timeout % should be the timeout ratio in the last 3 months?
I think it is a good idea.
I had the same pb as Mekk a few month ago: I entered a 1 day no vacation tournament , and I timeouted all my games because I missunderstood the starting date of the tournament : it began a few days earlier as I thought and I dind't connect the day the tournament began.
I timeouted 8 games in this tournament (those 8 games are my only timeouts here on chess.com) and since then I can't enter any tournament (and I won't for a long time).
Of couse it was my mistake but ... it is a bit harsh

I'd be keen to see a time limit on time outs.
The only two times I refuse a challenge is when I have too many games going or when the challenger has too many timeouts.
As Mekk pointed out, there can be reasons for this and t playing 1-day no vacation tournaments increases the risk.
A three month cut-off may be a bit short; what about six months?

I'm not yet a full member but am thinking about it.
I'm in the same scenario with the high timeout but have no excuse for it: in the early days I was just "checking out chess.com"... then lately I forgot to put my status on "Vacation". I was starting to think of playing against a "Fake" account a few games just to get that 10%. I wonder if some are doing the same thing (or just re-creating a new account).
I like the "penalty" option that aansel and Mekk have described. Maybe the system can also determine if the timeouts happened at once or over time (i.e. if the user isn't really serious and keeps forgetting a few games here and there, or it might have happened over a weekend when he/she wasn't there) and build the "trust/risk factor" of the user accordintly so to determine the penalty.
So educating users to not drop out of games but not penalizing so much the ones that accidently timed out.

Not to be a grinch about it, but if you have 63% timeouts, I'm sorry, I really don't want you in my tournaments, no matter how long it's been.
Timeouts wreak major havoc on tournaments, creating a lot of unearned wins and messing up the standings. They give undue advantage to slower players, because fast players may be done with their games before their opponents disappear. They're the sole reason I make a lot of the tournaments I direct premium-only -- because I figure if someone is serious enough to pay money for the site, they're not just going to up and leave suddenly.

Not to be a grinch about it, but if you have 63% timeouts, I'm sorry, I really don't want you in my tournaments, no matter how long it's been.
You know the difference between 6 out of 10 and 600 out of 1000, or don't you?
Also, whom would you prefer:
- a player who timed out 25 games once, 6 months ago, but returned and finished 25 games since then (with score 50%)
- a player who drops 10 games every month, but plays a lot, so his score is just 5%
% simply and just is not a good measure here.

Best to create a new account and close this one if you want to wipe your stats clean.
Most chess sites strongly advise against creating duplicate accounts, IIRC chess.com also does this. I'd prefer to avoid it.

Timeouts wreak major havoc on tournaments, creating a lot of unearned wins and messing up the standings. They give undue advantage to slower players, because fast players may be done with their games before their opponents disappear.
There is a simple solution at least one competing site apply: in case some player times out noticeable part of his tournament games (don't remember the exact criteria, it can be 1/3 or so), his results are removed from the table and ignored.
In fact, exactly the same happens in non-chess contexts. For example, if soccer team withdraws from the league after playing less than half of the matches, results against it are cancelled.

Also, whom would you prefer?
Personally, I'd prefer players who don't time out at all.
A few opponents have timed out when I've had an overwhelming advantage because they apparently felt it was less humiliating to do that than to resign. Or, worse: to play on in a Damoclean position and wait for the sword to fall.
Of course, that's only happened a dozen or so times in more than 1100 games completed -- not much more than 1% of my games -- so I'm not going to complain even though, at the time, it was disappointing.
But full marks for the grammar.

And maybe of related note: chess.com shows some strong bias towards play quickly, play often, have very reliable internet connection type of players. Restrictive timeout policy is one side. Lack of Fischer clocks is another.
Worst side is the very unfortunate you can't go to the vacation when you are to move in any of your games rule, which makes it impossible to take emergency vacation when something unexpected happens. Well, I'd have single timed out game instead of 38 if this rule wasn't used (I had my mobile phone with email, when I got first notification I tried logging to chess.com from it, while the browser was too limited to let me make moves, I was able to visit the settings screen and try asking for vacation. And, I got the aforementioned message).

Best to create a new account and close this one if you want to wipe your stats clean.
Most chess sites strongly advise against creating duplicate accounts, IIRC chess.com also does this. I'd prefer to avoid it.
Yes, my comment was in response to remby2 who appeared to be considering creating a new account and using it for precisely the purpose that it is disallowed (if you're going to manipulate your timout percentage, why not your rating while you're at it...). I was just suggesting an alternative that would be more in line with the intent of the rule.

And maybe of related note: chess.com shows some strong bias towards play quickly, play often, have very reliable internet connection type of players. Restrictive timeout policy is one side. Lack of Fischer clocks is another.
Worst side is the very unfortunate you can't go to the vacation when you are to move in any of your games rule, which makes it impossible to take emergency vacation when something unexpected happens. Well, I'd have single timed out game instead of 38 if this rule wasn't used (I had my mobile phone with email, when I got first notification I tried logging to chess.com from it, while the browser was too limited to let me make moves, I was able to visit the settings screen and try asking for vacation. And, I got the aforementioned message).
Better clock options are a whole other issue and I fully support them -- particularly those that provide more flexibility from move to move and more certainty with respect to the overall length of a game. I think that the lack of these is actually the root cause of a lot of the tournament complaints you see in these forums.
The vacation rule for non-premium members is in my mind fundamentally flawed and incredibly unfair as a result. If you have enough games on the go and a fast playing opponent is online (or someone with a long line of conditional moves) you could conceivably never be able to clear your games to turn your vacation on. I've proposed this a couple of times now, but I think you should be able to turn your vacation on for all games in which it is not your move, and then proceed to clear the games in which it is by making a move in each of them to switch each of them to vacation mode as well.
Full agreement on both points.

Best to create a new account and close this one if you want to wipe your stats clean.
Most chess sites strongly advise against creating duplicate accounts, IIRC chess.com also does this. I'd prefer to avoid it.
Yes, my comment was in response to remby2 who appeared to be considering creating a new account and using it for precisely the purpose that it is disallowed (if you're going to manipulate your timout percentage, why not your rating while you're at it...). I was just suggesting an alternative that would be more in line with the intent of the rule.
Yes I agree, it wouldn't be honest at all. I wanted to show how the % of timeouts could potentially affect players and could lead to cheating habbits if something is blocking them. I'm sure I'm not alone and do not want to encourage this type of behaviour.
I timed out 38 games in August. Leaving actual reasons apart (*) we can do simple math: as most tournaments (including those official) require timeout ratio < 10%, I need to finish 343 games before I am able to join one of those (or create my own).
Not very likely.
Timing out is evil, but current implementation makes anybody who happened to timeout many games for whatever reason (being out of the net for a few days would suffice) permanently excluded. In fact, creating new account looks like the only realistic option.
(*) I took chess.com vacation and left for my holiday. For some reason my vacation was not started or was cancelled, and most games I was playing timed out. I reported it earlier, unfortunately got no feedback.