Time Systems / Discussion

Discussion moved from Overtime:

(Sebastian:) RP proposed the following distinction between main time and overtime: [...]

Robert Pauli: Well, I rather proposed a new page content. I took care to keep both meanings above, not?

(Sebastian:) The original page contained two explicit definitions. I'm not a specialist on time systems, but it seems that these two adequately represented the current common usage of the term.

Your changed text does not define the term, but refers to it obliquely and hides the fact that it is used in two ways. Maybe the different meanings are still in your text. But it certainly is clearer to define them explicitly like any good encyclopedia would. You are invited to promote your view on a wiki wiki web; but you should keep an NPOV and not simply purge contradicting views.

This is best illustrated with a central statement of your concept: “Being in byoyomi not necessarily is being in overtime! Under Japanese Timing, for instance, there is no overtime, but certainly byoyomi.” This contradicts current usage. Many people define Japanese Timing as m minutes main time + n * b minutes/move byoyomi. As I see it, byoyomi is clearly distinguished from the main time. ^[1]

That is not to say that it isn't an interesting idea. Please don't get me wrong. I actually like the thrust of your contribution. It is a very elegant systematic approach. I see this concept as your invention, because for me, at least, the term "Canadian Time" as opposed to "Canadian Overtime" is new. I also like the fact that you created a page with a generic term for both: Round Down Timing.

The terminology still has some problems, which have been there before your changes, but are compounded by them.

[101] The term “time” can have at least 3 meanings, distinguished by careful writers with a modifier, which you usually leave out:

1. synonymous to “main time” (as in “Players are in overtime if their time runs out”)
2. synonymous to “time system” (as in your request to rename “JapaneseTitleMatchTimeSystem” to “Japanese Time”)
3. a general term for any time, including main time and overtime (as in: “Time Systems limit the period of time available to each player.” or “a game played with time limits”)

[102] For the term “overtime”, we also have three meanings:

1. the “second lives” of a player: 1st overtime = 2nd life, 2nd overtime = 3rd life and so on. ^[2] (could be called “overtime period”.)
2. a time system which is used after the first time system expired. (Thus, Canadian Time can be used as an overtime. This includes 1st through nth overtimes [meaning 1].). (could be called “time subsystem” or avoided altogether.)
3. a time system that switches from Absolute Timing to another time system (This includes main time as well as 1st through nth overtimes [meaning 1]; as in "Canadian Overtime [meaning 3] is a time system [that uses Absolute Timing as its main time and then Canadian Time? as its] overtime [meaning 2]" - see CanadianOvertime.) (could be called “overtime system”.)

Robert Pauli: Thanks for your love to detail, Sebastian (happen to have the same problem :--).

• Byoyomi: If many people think of "Japanese Time as M minutes main time + N * (B minutes/move) byoyomi", then I almost agree:
  it's Absolute Time (M) followed by Japanese Time (each move in B minutes, with N lifes).
• Invention: I'm really not trying to invent much, I'm trying to make things plain clear, kicking only what doesn't fit into the big picture.
• Canadian Overtime: If all but one are called "T Time", then the one standing out has to step back. So, what's Canadian Overtime then? Well, as the name suggests: overtime managed by Canadian Time. Put them as alias (not correct, OC).
• Tiled Time: It was described on the byoyomi page (glad Unkx80 did the trimming -- you'd cruzify me ;--) and was very worth to be kept. Don't know if it's applied, however, and native speakers please tell us if the qualifyer really suggests what it intends: being cut up in pieces of equal size? Could also be Parceled Time . . .
• Time:
  1. Maintime is just the time managed by the first member of a string of time systems, e.g. we're not really playing under Canadian overtime, but under absolute (main)time follwed by Canadian (over)time.
  2. Wouldn't object at all if someone renames each "T Time" to "T Time System" -- just didn't dare to be so wordy.
  3. We don't have to qualify "time"s normal meaning.
• Overtime:
  1. Don't confuse internal lifes used by a time system with overtime. 8 hours Japanese time in minutes can be seen as one move per minute plus 8 x 60 = 480 lifes, but I wouldn't speak of them as overtimes. So, no, there's no waterproof distinction.
  2. It's no time system, but time managed by a (sub) time system not initially in charge.
  3. This overloading just was my concession to what I found (page Canadian Overtime). Don't really like it. It's better to be wordy and name all parts of a time system.

Robert Pauli: Worked over the whole "mess":

• Fischer Time => Fischer Timing, etc.
• Tiled Time => Periodwise Timing
• "period" instead "unit" or "tile"
• Japanese Byoyomi (deprecated)
  • = Japanese Overtime
  • = overtime managed by Japanese Timing
  • = (loosely) Absolute Timing + Japanese Timing
• Canadian Byoyomi (deprecated)
  • = Canadian Overtime
  • = overtime managed by Canadian Timing
  • = (loosely) Absolute Timing + Canadian Timing

Time / Time System / Timing

Bob Myers: Whatever else one can say about the effort to restructure the pages about time systems in go, I think the choice of the word "timing" to refer to time systems is unfortunate. I've never heard this word used to refer to time systems. What is wrong with "time system", which after all is the name of this page? If people don't like that, "time management system" would still be better, or even "time control system". Personally, I prefer just "time".

SL is a fine place to introduce new terminology, such as a new word for a heretofore unnamed shape. Or to propose new terminology, marking it as a proposal. But first and foremost it needs to present current practice IMHO. In that sense as well, "time" or "time system" seems a better choice.

(Sebastian:) There's nothing wrong with "time system" - in fact, it's my favorite. But I have to admit that it is a bit lengthy, especially in terms like "Japanese Time System". I thought it was a nice idea to consistently use "timing" as an abbreviation for "time system", because it hasn't been used for any other meaning yet. I think the criterium should be: What is least likely to cause confusion? So I really don't think we should promote "time" for "time system".

Robert Pauli: Absolutely agree, Bob. "X Time System" would be way more systematic and crystal clear. I just feared, most would find it too clumsy. However, if most think as you, I'll obey -- no problem.

Overtime

Robert Pauli: OK, Sebastian, let's get over Overtime :--) I let you do it, humbly suggesting something like this:

In a game played with time limits, the time (if any) available after main time has been used up is referred to as overtime. One could also say that it starts when one (sub) time system ends and another starts.

I intentionally didn't mention "overtime" used as "time system" because, as far I see, this only is the case when qualified, and I covered that in Canadian Overtime and Japanese Overtime. Fine?

Now, whatever you do, please bring Overtime to a conclusion. (Don't want to see my name hanging around there forever :--)

(Sebastian:) Your efforts in cleaning up the mess we had in this area certainly deserve support, and I won't let you hang in the middle of the project!

But I need to know what your plans for Overtime are. What do you call the time system used in Japanese Title matches? My impression was that you used to call it "Japanese Overtime"^[3], but your table now says that this is only the name for the overtime itself [meaning 2]^[101]. Similarly, you seem to have changed what you used to call "Canadian Overtime [meaning 3]^[102]" to "Canadian Timing".[ This does not fix the confusion between overtime meaning 2 and overtime meaning 3[#102] but only shifts it from one word to the other, as Bob points out in Canadian Timing/ Discussion. Please take an earnest look at my lists^[101][102] and propose a unique term for each of the different meanings. (I just added some terms as proposals, but I really don't care too much what we call them, as long as we're consistent and not too much in conflict with existing usage.) I don't think "time managed by a (sub) time system not initially in charge" would fly as a term, because it's wordy and you don't use it yourself in the case Bob describes. (Please also see my e-mail which I just sent.)

Robert Pauli: To your questions, Sebastian:

• Japanese title matches use Japanese Timing: no Absolute Timing in front of it!
• Overtime really just is the time when you're over (main) time (or dead once).
• X (over)time is a kind of (over)time, the kind that's happens to be managed by time system X Timing, just like green apples are a certain sort of apples.
• X Timing is a kind of time system.

Let me try your 101 and 102 anyway, even if I might not get'em (intentionally not reading your mail, otherwise you'll never learn):

• 101 (Time)
  • Main time is called "main time", but can be managed by whatever time system -- even if that normally is Absolute Timing (= what you spent is what you pay).
  • Time system X is called "X Timing".
  • Hmm, don't know what to do with that one . . .
• 102 (Overtime)
  • Yes, overtime is one's second chance (qualifying it with "first", "second", etc. only makes sense if more than two time systems are combined).
  • No, overtime is not a (sub) time system.
  • No, overtime is neither a (combined) time system. However, I mentioned this use in Canadian Overtime and Japanese Overtime, and it's not too bad because "overtime" implies some "main time" and, if not expressed otherwise, main time is managed in an absolute fashion.

Chris Hayashida: I know I posted this somewhere else, but it seems applicable here. Can we get rid of the "Absolute Time followed by x Overtime?" I think if we define "Main Time" it might elegantly define most time systems. Absolute time has a main time period and no overtime. Japanese Time System has main time plus x y-second overtime periods (i.e. for KGS. You can put a footnote on the Japanese Time System page for title matches. Besides, isn't the Meijin just 0 minutes main time with 480 one-minute overtime periods? :) Canadian Time System has main time plus x minutes for every y moves. Seems like adding a definition for "Main Time" and not using "Absolute Time" with two meanings would make things less confusing.

Robert Pauli: I'm against superfluously adding main time to every time system.

It's way more elegant to to say

• Meijin is run by Japanese Timing than
• Meijin is run by Japanese Timing with zero main time.

(Yes, Chris, Japanese Timing of 8 h in 1 min periods is zero main time plus 480 periods of one stone in a min.)

It's also way more elegant to say

• The time system used for overtime is Canadian Timing than
• The time system used for overtime is Canadian Timing with zero main time.

Chris Hayashida: I think it is far more elegant to use:

• Canadian Timing uses x minutes main time with y minutes per z stones Canadian overtime.

Having no name for the frequent combination Absolute plus Canadian Timing is less a problem to me since we have to specify both parts anyway.

Chris Hayashida: I'd think that referring to the above time system as "Canadian timing." After all the main time and the overtime are subsets of the time system, right?

Absolute Timing is not used in two meanings, Chris. It's the way your supply on time can be deducted, the one where each second you take falls off your clock. Please notice that it's no longer Absolute Time but Absolute Timing.

Chris Hayashida: I think the problem I have with using the phrase in this way is everywhere I have seen "absolute time," it refers to sudden death. I have never seen the "thinking time"/"free time"/"main time" ever referred to as "absolute time" or "absolute timing" unless there was no overtime. For clarity's sake, I beg you to get rid of this usage. It's clearer to coin a separate term (one that's already in use, I might add) than it is to try and create a new meaning for an old one. It's more confusing the misuse of byouyomi.

It seems like your only gripe is that Japanese title matches do not fall neatly into this category. Since most of the deshi on Sensei's won't be playing in Japanese title matches, does it really matter? I think title match time systems can just be described separately.

rubilia: Sorry Robert, I don't see much use of redefining established terms here, either - no matter if the proposed meanings seem more logical. Just assume 10 % of all Go players stumble across SL from time to time, and some of them may even read your time system pages. Now, there are two distinct things, a and b, usually called A (main time) and B (absolute time). If you start to use B as reference to a, and even claim the authority to define which meaning has to be regarded as correct, do you really expect to change the commonly used language by that, or at least, to make things less confusing? Honestly, to me, it doesn't look like those hopes are justified. (Btw, the established designations CAN be used in a rather systematical way, as well - see below.) ----

[1] See e.g. ChronosManual or Benjamin Teuber / Diary in Japan. You may know better, but I realy doubt that in the Japanese Time system "players get a number of time units of same size", as you say. The main time is not the same size as the byoyomi periods, and AFAIK it is not chopped up into tiles. At least I haven’t noticed that yet.

[2] This only became clear to me now. This meaning was not mentioned in the original list although it has been used before. Maybe it doesn't need to because it is only used with the ordinal number.

[3] There was beauty in your original concept of dividing a time system into subsystems. But it's no use if it's only applied half-heardedly. This is actually why I wrote that it was a proposal - this way we can keep it crystal clear, free from current unsystematic usage of terminology. I don't mind taking your name off that page, but I think you deserve credit for the concept.

Bob Myers: In general, and with all due respect to RP's efforts in thinking about and restructuring the time/timing/time system pages, I find something missing. I think there is an issue here related to the objective of these pages or even of SL itself. I think we should remember that SL is first and foremost a resource for average players.

When we start talking about "piping" time systems, we are outside the realm of explaining things to the average player. These pages have now moved into the realm of an attempt to do a theoretical deconstruction of the inner structure of time systems, not only using new and unfamiliar vocabulary but even proscribing the use of terms that everyone knows and uses.

To me, it seems that the *either* a new "friendly" page with an overview of timing sytems from the human standpoint should be written (I don't really know anything about time systems myself and can't contribute directly), *or* the current pages should be prefixed/bundled into a section about a proposed logical analysis of time systems for those who are interested.

Such a friendly page would tell the player of the two main systems in practical use: tournament time control (main+byoyomi) and net time control (main+Canadian). I know that there is logical fuzziness here. My point is that the description needs to relate more closely to what people experience as they play go daily on the net or in tournaments.

Totally as an aside, the main page says that Fischer timing is one of the common systems, but I've played go for way too long and never even heard of it. Is this a European thing?

Robert Pauli: You don't believe that someone doing Go can cope with this terminology? Come on.

The idea of piping time systems isn't really new, Bob. Indeed, we're doing it. Your flag falls, but, hey, you don't lose on time, you continue -- now with changed restrictions. As restrictions change, do the (sub) time systems.

Even you make parts. What's the big difference between

• main+byoyomi / main+Canadian, as you put it (or did I infect you already ;-), and
• Absolute plus Japanese Timing / Absolute plus Canadian Timing,

as I humbly suggest?

BTW, I never hearded of Absolute Time(ing) before, it always was Sudden Death to me, but I kept it. It's a good name! I can learn. Is it only me?

Chris Hayashida: The problem is that this usage of Absolute Time/Timing goes against other definitions. It's not that I don't want to learn. But I also want to make this page something I can refer to beginners to before a tournament.

Anyway, I (hopefully) now put it more plain on Time Systems.

Fisher Timing was listed as common before I started. I neither believe that it's used much with Go, anywhere: changed it.

Now I'll calmly sit back and watch how things get muddled again -- I had my try. :--)

Chris Hayashida: Just for reference:

KGS uses the following names for time systems (with names used for settings in parenthesis):

• None
• Absolute (main time only)
• Byo-yomi (main time, byo-yomi time, byo-yomi periods)
• Canadian (main time, byo-yomi time, stones per byo-yomi)

IGS has the following notes in help time:

• time (per player, but does not specify "main" or otherwise)

The Panda-gGo client refers to this time as "main time."

The IGS help byoyomi command uses the following names:

• byoyomi (byoyomi minutes, with the 25 moves as the default per byoyomi period)

In help stats the number of stones per overtime period is referred to as byo-yomi stones. The Panda-gGo client refers to this time as "byo-yomi time."

I didn't have the time to look at other IGS clients, or published rulesets for Ing or the Go Congress.

Bob Myers: RP, thanks for your answers. But I continue to feel that these pages as currently structured do not answer the needs for a readable, understandable, accurate overview of time systems used with go.

On the second paragraph of this page, we immediately are told that "as soon as players exhaust their resources...", with a footnote we need to follow to go down and find out that this unfamiliar word (in the time context) is referring to "time and/or lives", where "life" is another unfamiliar term. not referred to or defined anywhere else on the page. Then we have the caveat about "no matter if their turn was complete", which is a legalistic rules issue (I assume this refers to the case where the player forgets to hit the clock).

Moving on to the next paragraph, we find that the first common system is "Absolute plus Canadian". The word "absolute" here is absolutely wrong. Absolute implies that there is no recovery from exhausting it. I do not know about the nuances of cognates of this word in other European languages but in English it doesn't work, as Chris H has repeatedly tried to point out. I completely fail to understand why we can't call this what it is, using the word always used for it, which is main time. The problem may be that you're looking for a "name" for this "timing system", that can be combined algebraically with other timing systems, whereas actually it's just, ummmh, main time.

Even in your A+B taxonomy of timing systems, there is only one timing "system" that does, or could, come first--what you call "absolute". So why go to all the trouble of setting up some kind of set of theoretical set of building blocks for time systems that can be combined in all different ways. This is a bit of an overplay. And initial time is actually an aspect of virtually all time systems, other than the Japanese pro style. So the easiest approach is simply to assume that all timing systems come with the initial time. Voila. Canadian means initial time with Canadian overtime. Japanese means initial time with Japanese overtime. Then we simply need to note that Japanese professional games are played with no initial time.

To summarize, in my mind the top-level "Timing Systems" page should propose the following basic time systems:

• Amateur tournament time style ("Japanese"), involving an initial time allocation (main time) with time thereafter handled in the Japanese style (play each stone within x seconds, with the ability to spend an additional y seconds thinking no more than z times)

• Net-style time ("Canadian"), usually involving an initial time allocation (main time) and Canadian overtime (x stones in y minutes). But I don't think this needs to be called "Main+Canadian" (certainly not "Absolute+Canadian"). As I mentioned above, there is no reason to deprecate the name "Canadian time" for the entire system, since as Chris Hayashida points out that's what KGS calls it. If someone really wants to describe a Canadian time system with no main time, then they could call it, ummm, "Canadian time with no main time".

• Japanese professional tournament style, which uses the Japanese style for the entire game, with no initial block of time.

You'll note in the above, I've tried to describe the major time systems really briefly. I think this is important in the sense that it allows someone to visit this one page and get real information, however basic, on actual time systems.

Some of the comments below should rightfully go in the individual discussion pages but I've chosen to keep them together here.

On the "Canadian Timing" page, we have a first sentence which says "Canadian Timing is a time system that manages time in a periodwise manner". This violates the writing rule that says to say what you want to say. It should say, "In Canadian Timing, the player has to play x stones within y seconds." Since we are now using Canadian to refer to the whole system, we might want to add: "after using up his initial block of time".

On the Japanese Timing page, we have the opaque three-point summary I criticized earlier, which should simply say, "Players have x seconds to play each stone; but they may take an additional y seconds of thinking time a maximum of z times during the game." It should then tie back to the overall timing discussion by pointing out that this approach is often used for overtime (yes, it *is* overtime, by common usage) in amateur tournaments, and used for the entire game by Japanese professionals. All the stuff about professional games should be moved to a new page called Time Management in Japanese Professional Games?.

In any case, the current Japanese Timing page is just wrong in one respect: that, in its terminology, "all the time periods are of the same size". For instance, in the NHK Senshuken games, each stone must be played within thirty seconds, but the additional thinking periods are one minute each.

Finally, on the Time Systems page the "Less common" systems should IMO be moved to a Obscure Timing Systems? page referenced under "See also".

This has been my two cents. I appreciate everyone's work in the past and future trying to make these pages as good as possible.

Robert Pauli: Chris, Bob:

I know exactly what you mean and I agree that practical relavance isn't yet met. Please be patient. I'll work on it (if this discussion doesn't suck up all my time). Of course, you're welcome to participate ("In practice this time system is dealt with as follows . . .").

I'm admittedly stressing systematics (looking beyond particular time systems), precision (Stone hits board . . . see below), and elegance (the darn "plus"). You're stressing touch with real life. OK. How about having both? Shouldn't be impossible.

Special points:

• Thanks for the "real" names, Chris. I'll think about (named) parameters . . .
• "Resources" might be unfamilar, but not incomprehensible, Bob.
• "Complete": Stone hits board, then flag falls: lost? Yes!
• "Absolute": Again, Bob, not my invention. "Absolutely" wrong? Not at all! It deducts used time in an absolute manner. Please don't argue for result, Bob.
• "No recovery from exhausting [resources]" is part of every time system, Bob, not specific to Absolute Timing at all. If your resources are used up, you're gone (no matter if your stone hit the board or not).
• "Main Time": What's main time, Bob, when, for instance, we're doing Fischer Timing? What if a tournament is run with Fischer plus Japanese Timing? Where's your main time then? Your claim that only Absolute Timing can be used as front end simply is wrong.
  • [5]rubilia: I don't see any serious problem there. Some examples, and how they can be called systematically using well-known designations:

1. Absolute time: main time only, e. g. "30 min abs".
2. Japanese time system: e. g. 30 min main time, and subsequently, 3 * 30 s Japanese overtime
3. Japanese Professional: the same, but with zero main time (e. g. 480 * 60 s Japanese overtime)
4. Bronstein time system: like 2., but without forced subsequence
5. Canadian time system: e. g. 30 min main time, and subsequently, 5 min per 15 stones Canadian overtime
6. Fisher time system: main time e. g. 20 min, Fisher overtime 10 s per move (no forced subsequence)
7. hourglass time system: e. g. 5 min main time, hourglass overtime (no forced subsequence)
8. customized XYZ time system "Fischer plus Japanese" (if that really happens): e. g. 20 min main time, and Fisher overtime 10 s per move (no forced subsequence); subsequently 3 * 30 s Japanese overtime.

• Sure, Bob, Chris, we could incorporate main time into every time system, yes, but it would be utterly unelegant -- if you have any sense for that. Should I also put it into Fischer Timing? into Bronstein Timing? into Hourglass Timing?. If piping already is a fact, why not build upon it? Don't be so narrow minded -- think big.

Chris Hayashida: I am thinking big. Instead of trying to get all of the time systems to conform to a mold, I'm trying to get descriptions that are easy for beginners to read. From what I have read, the three pages you refer to above are clear enough. But the "Absolute plus..." on the Time Systems page for Canadian and Japanese Time Systems is not clear. It made more sense when it was three pages: Japanese Professional Time Systems, Japanese Time Systems, and Canadian Time Systems. Trying to squeeze Japanese Time Systems into the same page with Japanese Professional Time Systems seems to have caused some of the confusion. Again, from the reader's point of view, it's not helpful. At most, maybe a footnote on the Japanese Time Systems page saying "Professional games are managed differently" and a pointer to another page would suffice.

Bob Myers: I've got an idea. Let's personalize this and start impugning each other's sense of "elegance" or whatever else. Seriously, your use of the term "elegance" indicates the problem here. You are trying to write elegant pages. Chris and I are trying to get to pages which are friendly, understandable, and relate to what go players experience every day. Above, you've shown you understand the problem by acknowledging that you're "stressing sytematics" and that the pages as they stand "lack practical relevance", but then go on to reject all the proposals oriented towards practical relevance. It would be unfortunate if all this resulted in the necessity to have separate pages, one called "A Friendly Overview of Time Systems" and another called "A Theory of Time Systems and their Piped Components."

• What you, Bob, can't avoid to call "Japanese Style" (smile) is what I call "Japanese Timing" -- big deal. It's out there, we're not hallucinating. It deserves a name.
• "NHK Senshuken": Anybody please give me a better description of the time system used there (didn't get when overtime starts, Bob). I'll try to pipe it together -- or make up the missing part . . .
• "Obscure Timing Systems" is an example of you're narrow-mindedness, Bob -- sorry. I'll certainly not do that. Your "writing rule" with regard to Canadian Timing neither shows you very consequent. Suddently you feel the need to come to the point, only after having rejected this for Japanese Timing (the definition there is crystal clear).

Bob Myers: In the NHK Cup, starting at the beginning of the game, players have 30 seconds to make each move, but have 10 extended thinking periods (kouryo jikan=考慮時間) of one minute. (For those interested in "real-world" implementations of time systems, the above is essentially a word-for-word translation of how the time system is described at the beginning of each program.)

And for those interested in arcane variations of byo-yomi conventions, and IIRC, the seconds are read off at the 10, 20, 25, and 28 second point of each thirty-second period, and the 30, 40, 50, 55, 58 second point of each extended thinking period. However, after all ten extended thinking periods have been exhausted, the last ten seconds are read off one by one: ichi, ni, san, with the player in question losing when the byo-yomi-gakari pronounces the dreaded "juu".

Robert Pauli: So, Bob, this would be Japanese plus Japanese Timing:

• the first with one period of 30 seconds
• the second with 10 periods of each one minute

See, "my" system works. I'll address the rest later, but, yes, I neither want two sorts of pages . . .
(BTW, are you the Endgame Power Bob -- happen to have Go Review Wi'77 and Sp'77? Great stuff!)

Bob Myers: Yes, that's me. On another note, I think pages such as Steady Average Timing should include notes to the effect that these time systems are mere theoretical artifacts which never have been and never will be used in real life, to avoid confusing average players.

Robert Pauli: Nice to meet you. Agree generally, but in this special case it indeed was (is?) used, therefore the alias: Milton Keynes Timing (see BGA Tournament Rules).

---

Jef Pearlman (puk on KGS): Sorry to jump in... I scanned this and other pages, but wasn't able to find exactly what I'm looking for. Has anyone proposed a hybrid system like this? Start with an optional "main time". Then, each move gets x number of seconds, plus a Canadian-style y moves in z seconds "overflow."

For instance, you could say each player gets 5 (x) seconds per move, plus 1 (z) minute per 25 (y) moves. If you make a move within the initial x seconds, your count of remaining moves in that overtime period goes down, but you still have 1 full minute overflow. If you go over the 5 seconds, you eat into your 1 minute. You only run out of time if you go through your 5 seconds and your entire pool of overflow time. At the end of each y moves, the overflow time (z) is reset to full. In this example, is still roughly 3 minutes maximum per 25 moves (compared to Canadian 3/25).

The benefit would be that 1) you can spend extra time on certain moves, yet 2) no matter how badly you manage your time, you will always have at least 5 seconds per move. I like to play blitz, but in reality, I don't want to win on time, and this deals with my major complaints about the time systems used for blitz. Canadian forces you to do a little too much long-term time management, and if you mess up, you can dig yourself into a hole there's no way out of (and your opponent can force you to make stupid plays during the last few moves of the time period). In normal byo-yomi, though, there's very little ability to allocate longer periods of time where necessary to think through a move.

Please let me know if this was the wrong place to put this, or if this has been dealt with already, and feel free to erase it.