about the chinese readings of the kanjis

[juda]

After studying kanjis for a while now, i have noticed that some pronunciations are always together
For example a kanji that has the pronunciation sai also has sei as a second. The same goes for sei and shou, mei and myo, bin and ben, kon and kin...
My question is, can i say that all kanji that has X reading will have Y as second reading ?

[Disco]

I dont think that thats necessarily true. A word can have multiple second kanji. Ex: Sai
再生saisei、再現saigen、再会saikai、最高saikou、最強saikyou、最低saitei. So to answer your question; No, one cannot assume that is a kanji has X reading, that Y will have a certain reading. Hope I shed some light on that situation ; )

[Hyperworm]

I don't think the question is whether sai will always be followed by sei; the question is whether all kanji that can be read "sai" can also be read "sei" (=also appear in words where that kanji is read "sei").

e.g. 歳（さい） has 歳暮（せいぼ）

It's an interesting question, and may be true for a limited few readings, but I think I can find counterexamples for each of the ones in the OP.
sai->sei : 再 has no words where it's read せい
bin->ben : 敏 has no words where it's read べん
kon->kin : 根
sei->shou : 製
mei->myou : 鳴

[Disco]

I see, so like 今年　and 来年？ same position but different pronounciation. toshi/nen

[juda]

I see, so i can say that it will Y as second pronunciation but i might never see it. I'm ok with that. But can i say, that if a kanji has X pronunciation the second can only be Y?

@disco, no that is little different toshi is the Japanese reading while nen is the chinese one, while the examples given here are all chinese readings. also as far as i know you can't determine the reading of a kanji by its position in a word.

[blutorange]

Sampling the set K of kanji as used in Japan, there are certain pairs of (perhaps not official) on-readings (A,B) so that one can observe a statistically significant correlation between a certain kanji possesing the reading A and B. That correlation, however, is not 100%, therefore the statement
for all on-readings (A,B)∈O, for all kanji X∈K: A is reading of X ⇒ B is reading of X
(O being the set of on-reading (A,B) with the property that there exists a kanji X∈K so that A and B are readings of X.)
is not true.

[AJBryant]

I see, so like 今年　and 来年？ same position but different pronounciation. toshi/nen

No, like how 正 (among many other kanji) can be either SEI or SHÔ, and 名 (among many other kanji) can be MEI or MYÔ, etc.

The thing is, so-called on-yomi is actually several different readings that can be specifically identified with a Chinese dynasty or state. Certain words and compounds came into Japanese from China at different times, and the time these words came in can be identified by whether the word has the Tô-on (Tang reading) or Kan-on (Han reading), and so on.

This section of the WIkipedia article on kanjiwill provide more information on the subject.