SH Goal Filters

[A Former User]

@richard-latimer Looks like a decent start for the filter. I would be conscious of reading too much into individual scorelines too early though, if you have only 100 results then some of the scores may only have 5-10 results which is too small a data set to draw much from. I am quite cautious of taking too much from scores where i only have 50 or so results, im just using the results to build a picture

[Richard Latimer]

@richard-latimer said in SH Goal Filters:

@lee-woodman said in SH Goal Filters:

@richard-latimer do you have an example filter?

My one and only filter at the moment is quite basic but it's doing the job. I start with 3+goals and then I add 1.6-2.3 ave home goals scored. Any more and the results go downhill. Any less and it's the same.

After that I make minor alterations based on odds, ave H1/H2 goals etc for both fhg and SHG.

When drawing 0-0/1-1 and no H2 goals has proved best but limits selections. Approaching 100 selections and close to 90% 60mins+ and 85% 70 mins+

Hence I'm now toying around with other scores as well.

[Richard Latimer]

@lee-woodman said in SH Goal Filters:

@richard-latimer do you have an example filter?

My one and only filter at the moment is quite basic but it's doing the job. I start with 3+goals and then I add 1.6-2.3 ave home goals scored. Any more and the results go downhill. Any less and it's the same.

After that I make minor alterations based on odds, ave H1/H2 goals etc for both fhg and SHG.

[A Former User]

@richard-latimer do you have an example filter?

[Richard Latimer]

@lee-woodman said in SH Goal Filters:

@richard-latimer Those figures are irrespective of what else has happened already in the half. I’ll look later whether it makes a difference and that might help some way towards my ‘information hole’ in that the HT score may not equal the entry score, however, every other time I’ve checked it’s not made a noticeable difference to the figures

It's weird because the stuff I'm looking at seems to make a massive difference.

[A Former User]

@richard-latimer Those figures are irrespective of what else has happened already in the half. I’ll look later whether it makes a difference and that might help some way towards my ‘information hole’ in that the HT score may not equal the entry score, however, every other time I’ve checked it’s not made a noticeable difference to the figures

[Richard Latimer]

@lee-woodman said in SH Goal Filters:

@finn-kristensen I tend to do similar with regards to the parameters in the filter. So whilst my average overall goals is 3.0, i do record as low as 2.5, so if i get further information like these scorelines then ill have a quick check and see if 3.0 is still ideal. Maybe now i have additional information it can be set to 2.8

Most of my filters are set up with additional criteria as an option so i may have say 3 main criteria but then ill add more criteria just set to >= 1 then i can check later to see if that criteria adds anything of value. And most of my filters are set up from 'master' filters with say 20 wide criteria which then can be analysed later to see which figures matter and what figures gives a good balance

I've started just messing around with one base filter and just analysing all the different angles.

[Richard Latimer]

@lee-woodman said in SH Goal Filters:

@darri 70 or 75' jumped out to me aswell, i like the odds range and trading around evens at (hopefully ongoing) 60%+ SR feels a good risk/reward ratio with reasonably low variance

Whichever way i look at it, 1-0 and particularly 0-1 just doesnt seem to be able to work. The figures i quoted are based on HT scorelines so ive looked to see if it makes a difference if theres a goal 46-70/75 and it doesnt. I appreciate its not just SR thats important but from my experience these 2 scorelines dont see a big enough reduction in odds to justify the drop-off in SR.

The filter itself tends to have a strong home team based on scoring goals (as opposed to being based on odds, ELO, league position etc) so whilst the home isnt always favourite they tend to be able to hold their own on the basis of scoring but generally both teams are solid scoring teams with an element of conceding their fair share. Ive got a filter which leans towards a stronger home team and one that leans towards a stronger away and the pattern is the same, so whilst im tempted to start recording odds & favourites i have to be conscious of recording too much due to time. I can pull in the ELO ratings but not sure whether a stronger ELO rating is indicative of stronger favourite

Ill have a think today about what else i can add in to either assist or backup the information, usually after posting something like this and reading feedback gets my brain going!

Are the figures quoted for first second half goal or does it not matter how many have been scored prior?

[A Former User]

@finn-kristensen I tend to do similar with regards to the parameters in the filter. So whilst my average overall goals is 3.0, i do record as low as 2.5, so if i get further information like these scorelines then ill have a quick check and see if 3.0 is still ideal. Maybe now i have additional information it can be set to 2.8

Most of my filters are set up with additional criteria as an option so i may have say 3 main criteria but then ill add more criteria just set to >= 1 then i can check later to see if that criteria adds anything of value. And most of my filters are set up from 'master' filters with say 20 wide criteria which then can be analysed later to see which figures matter and what figures gives a good balance

[A Former User]

@darri 70 or 75' jumped out to me aswell, i like the odds range and trading around evens at (hopefully ongoing) 60%+ SR feels a good risk/reward ratio with reasonably low variance

Whichever way i look at it, 1-0 and particularly 0-1 just doesnt seem to be able to work. The figures i quoted are based on HT scorelines so ive looked to see if it makes a difference if theres a goal 46-70/75 and it doesnt. I appreciate its not just SR thats important but from my experience these 2 scorelines dont see a big enough reduction in odds to justify the drop-off in SR.

The filter itself tends to have a strong home team based on scoring goals (as opposed to being based on odds, ELO, league position etc) so whilst the home isnt always favourite they tend to be able to hold their own on the basis of scoring but generally both teams are solid scoring teams with an element of conceding their fair share. Ive got a filter which leans towards a stronger home team and one that leans towards a stronger away and the pattern is the same, so whilst im tempted to start recording odds & favourites i have to be conscious of recording too much due to time. I can pull in the ELO ratings but not sure whether a stronger ELO rating is indicative of stronger favourite

Ill have a think today about what else i can add in to either assist or backup the information, usually after posting something like this and reading feedback gets my brain going!

[A Former User]

@lee-woodman hey bud, just read through this, superb analysis for me im only going to say how it would relate to my own knowledge (late goals) that change on the filter 2 in the +75 min range is huge. The general lcs being 63.5% is still profitable i can almost guarantee that mate. At 75mins your average odds will be around between 1.8-2.0 with heavy favs, so even higher if tighter games. However the change from 63.5% to 67.7% is massive given the odds youd be playing with. If that remains consistent, quite honestly you are going places. For 75 min onwards you only need around 57% to be break even, so your smashing it. Keep this up mate, you might find alot of people will be heavily looking into this regardless if its filter only type trading (me), terrific work on the analysis. While inplay watching and research will help these results too as you have pointed out in another post, id be happy making a passive income on this regardless. i see your now tracking even more stuff to try see if there is further improvement but this is a great baseline stat, any improvement towards 70% would seriously be impressive. On the 1-0/0-1 side of things knowing if that is the fav or dog that are ahead would be good data to know, might change the overall look of said scorelines

[Richard Latimer]

@finn-kristensen said in SH Goal Filters:

@richard-latimer said in SH Goal Filters:

@finn-kristensen said in SH Goal Filters:

@lee-woodman said in SH Goal Filters:

My regret on the data recording for the filters is that i didnt record the scores at time of entry (ie what happens if 1-1 at 60', or 0-1 at 70' and so on) and secondly who the favourite was, and how heavy favourites were they. Ive started to record some of that now but i think im going to do it across the board. Its hard to catch every bit of relevant data as its time consuming but hopefully itll be worth it

Excellent analysis - very interesting! Your second filter looks particularly good, especially when excluding the 1-0/0-1 HT scores.
I know the feeling about getting some ideas later when you already have a lot of data. What I do now is I always include as many variables as I can, but I don't include limits for them (say, if I include ELO, as I always do, I will just use ">= 1000" for both teams, which won't remove any relevant matches - I will then analyze them later to see if the difference in ELO has any significance for the filter). I also try to record various odds at kick-off, but as you say, it's very time consuming so a bit of a drag.

For odds try using ave bookie odds. Way easier to do as you can collect data after the fact if you miss the game.

Yeah, for odds I mostly use flashscore which shows odds for the most common bet markets from a number of bookies incl Bet365, which I typically use for reference.

oddsportal is my go to as it collates them all to give an average.

[Finn Kristensen]

@richard-latimer said in SH Goal Filters:

@finn-kristensen said in SH Goal Filters:

@lee-woodman said in SH Goal Filters:

My regret on the data recording for the filters is that i didnt record the scores at time of entry (ie what happens if 1-1 at 60', or 0-1 at 70' and so on) and secondly who the favourite was, and how heavy favourites were they. Ive started to record some of that now but i think im going to do it across the board. Its hard to catch every bit of relevant data as its time consuming but hopefully itll be worth it

Excellent analysis - very interesting! Your second filter looks particularly good, especially when excluding the 1-0/0-1 HT scores.
I know the feeling about getting some ideas later when you already have a lot of data. What I do now is I always include as many variables as I can, but I don't include limits for them (say, if I include ELO, as I always do, I will just use ">= 1000" for both teams, which won't remove any relevant matches - I will then analyze them later to see if the difference in ELO has any significance for the filter). I also try to record various odds at kick-off, but as you say, it's very time consuming so a bit of a drag.

For odds try using ave bookie odds. Way easier to do as you can collect data after the fact if you miss the game.

Yeah, for odds I mostly use flashscore which shows odds for the most common bet markets from a number of bookies incl Bet365, which I typically use for reference.

[Richard Latimer]

@finn-kristensen said in SH Goal Filters:

@lee-woodman said in SH Goal Filters:

My regret on the data recording for the filters is that i didnt record the scores at time of entry (ie what happens if 1-1 at 60', or 0-1 at 70' and so on) and secondly who the favourite was, and how heavy favourites were they. Ive started to record some of that now but i think im going to do it across the board. Its hard to catch every bit of relevant data as its time consuming but hopefully itll be worth it

Excellent analysis - very interesting! Your second filter looks particularly good, especially when excluding the 1-0/0-1 HT scores.
I know the feeling about getting some ideas later when you already have a lot of data. What I do now is I always include as many variables as I can, but I don't include limits for them (say, if I include ELO, as I always do, I will just use ">= 1000" for both teams, which won't remove any relevant matches - I will then analyze them later to see if the difference in ELO has any significance for the filter). I also try to record various odds at kick-off, but as you say, it's very time consuming so a bit of a drag.

For odds try using ave bookie odds. Way easier to do as you can collect data after the fact if you miss the game.

[Finn Kristensen]

@lee-woodman said in SH Goal Filters:

My regret on the data recording for the filters is that i didnt record the scores at time of entry (ie what happens if 1-1 at 60', or 0-1 at 70' and so on) and secondly who the favourite was, and how heavy favourites were they. Ive started to record some of that now but i think im going to do it across the board. Its hard to catch every bit of relevant data as its time consuming but hopefully itll be worth it

Excellent analysis - very interesting! Your second filter looks particularly good, especially when excluding the 1-0/0-1 HT scores.
I know the feeling about getting some ideas later when you already have a lot of data. What I do now is I always include as many variables as I can, but I don't include limits for them (say, if I include ELO, as I always do, I will just use ">= 1000" for both teams, which won't remove any relevant matches - I will then analyze them later to see if the difference in ELO has any significance for the filter). I also try to record various odds at kick-off, but as you say, it's very time consuming so a bit of a drag.

[A Former User]

My regret on the data recording for the filters is that i didnt record the scores at time of entry (ie what happens if 1-1 at 60', or 0-1 at 70' and so on) and secondly who the favourite was, and how heavy favourites were they. Ive started to record some of that now but i think im going to do it across the board. Its hard to catch every bit of relevant data as its time consuming but hopefully itll be worth it

[A Former User]

And this is the info for my first filter:

[A Former User]

As ive changed my trading a bit this last month i decided to have a deeper analysis of my 2 filters. I had a pre-conception that 0-0/1-1/1-0/0-1 were the better scores to trade as they indicated tighter games and therefore the game was still open but ive started to notice more that im getting a lot of losers from 1-0/0-1 HT scores and wanted to see how it translated from an anecdotal thought to actual data. Below is the data for my 2nd filter which is standing up really well statistically now its at nearly 350 games:

So ive broken it down into a few different score categories. Ive colour coded the SR for easy reference, Orange is neutral (within 2%) compared to the filter average, red is a drop-off of 2% or more, blue is an increase of 2% or more

Some of the scores and figures really jump out to me and whilst i dont plan on simply trading the data alone, it starts to build up a picture of how things are likely to play out and the best sorts of times to be getting involved

[Martin]

@lee-woodman said in SH Goal Filters:

I know filters are very much out of fashion at the minute but i thought id do an update of the 2 filters im doing, The 2nd filter is doing really really well. Ive updated the figures into the original post and these are the changes ive made:

Filter 1:
Removed Norwary OBOS
Changed scope to 5 games

Filter 2:
Removed both Norway leagues
Looked at Average scored by FT rather than home or away individually

Its interesting aswell that the HT score seems to have no relevance to the figures, so whether its a draw, a one goal margin or the game is all but won then the figures stay roughly the same

that's what I love about these that the scoreline doesn't seem to matter!

[Ryan]

@lee-woodman The filters will never be out of fashion everybody still using them anyways lol

HT score having no relevance is very interesting - especially if the game is all but won as well.