A general hello and help request

[Kevo00]

I had a quick look at the Gondborough game. What others have said above is basically correct. You have too few trains on too much infrastructure, which to me seems to be too widely spread around the map to manage. Many of the lines are very simplistic end to end lines, with no intermediate stations (even when they run past quite big towns) and only a few trains on them. Where the trains have up to date locos, the trains themselves are actually making money, but the amount of infra per train is massive. On a few lines, trains do not even share infra but have their own individual line, which is incredibly inefficient. I'm guessing most of the passenger lines in this company were inherited from the AI which you bought; this AI went bankrupt because its infra costs were too high and you simply inherited the problem.

The only way to save this company that I can see would be to close down most of the lines and massively increase frequency or train length on the electrified ones so that you got some sort of critical mass. I have never played the default set right through in OTTD but it looks like you have pretty much reached the end of diesel's viable life, so perhaps close the diesel lines altogether and move your resources to the electrified lines. Or load up something like the 2CC newgrf which will still have diesels you can use (not officially recommended but its never crashed a game for me) and close the electric lines because they will be costing more to maintain (non electrified lines are costing £9m a year, electrified £40m a year).

General tips:

1. Get the ratio of trains to infra as high as you can. Short trains are not necessarily a problem if you have a high service frequency - using MUs for example.
2. Use one way and path signalling to densen up service patterns.
3. Build a network, not a load of isolated end to end lines. Even for freight it is better to try and serve multiple industries with one line - look for areas with high industry density and target those. This helps make serving industries that are far apart viable. For passengers look for clusters or corridors of towns you can serve with intermediate stations.
4. Fully research an AI's activities and spread before you buy it - why is it going bankrupt? Due diligence, if you like.
5. Consider playing with CargoDist on because it will encourage network thinking.
6. I tend to prefer playing smaller maps on my own because then I can learn my way round the map and don't loose track of things too much. Maps based on real geography are even better for this.

Edit: I have spotted some other very weird stuff - there are lots of stations on their own with no service at all, and even some steam locos running on an electrified line! The station tiles are costing £10m a year to maintain. And one bit where a single track coal line runs parallel to a double track line. Management, management, management!

[Baldy's Boss]

I had a quick look at the Gondborough game. What others have said above is basically correct. You have too few trains on too much infrastructure, which to me seems to be too widely spread around the map to manage. Many of the lines are very simplistic end to end lines, with no intermediate stations (even when they run past quite big towns) and only a few trains on them. Where the trains have up to date locos, the trains themselves are actually making money, but the amount of infra per train is massive. On a few lines, trains do not even share infra but have their own individual line, which is incredibly inefficient. I'm guessing most of the passenger lines in this company were inherited from the AI which you bought; this AI went bankrupt because its infra costs were too high and you simply inherited the problem.

The only way to save this company that I can see would be to close down most of the lines and massively increase frequency or train length on the electrified ones so that you got some sort of critical mass. I have never played the default set right through in OTTD but it looks like you have pretty much reached the end of diesel's viable life, so perhaps close the diesel lines altogether and move your resources to the electrified lines. Or load up something like the 2CC newgrf which will still have diesels you can use (not officially recommended but its never crashed a game for me) and close the electric lines because they will be costing more to maintain (non electrified lines are costing £9m a year, electrified £40m a year).

General tips:

1. Get the ratio of trains to infra as high as you can. Short trains are not necessarily a problem if you have a high service frequency - using MUs for example.
2. Use one way and path signalling to densen up service patterns.
3. Build a network, not a load of isolated end to end lines. Even for freight it is better to try and serve multiple industries with one line - look for areas with high industry density and target those. This helps make serving industries that are far apart viable. For passengers look for clusters or corridors of towns you can serve with intermediate stations.
4. Fully research an AI's activities and spread before you buy it - why is it going bankrupt? Due diligence, if you like.
5. Consider playing with CargoDist on because it will encourage network thinking.
6. I tend to prefer playing smaller maps on my own because then I can learn my way round the map and don't loose track of things too much. Maps based on real geography are even better for this.

Edit: I have spotted some other very weird stuff - there are lots of stations on their own with no service at all, and even some steam locos running on an electrified line! The station tiles are costing £10m a year to maintain. And one bit where a single track coal line runs parallel to a double track line. Management, management, management!

Which save of Gondborough did you look at?...I demolished 12 orphan stations between the July 31 and August 26 saves.
The ONLY AI I bought out of bankruptcy,as I recall,was the one with train numbers starting about 255,in March 2003.As a rule I buy AIs that represent a competitive threat and have better profit margins than my own.Any diesel lines were acquired from AIs though the trains with numbers around 200 got new diesels to replace old ones in some cases...I prefer to go from steam straight to electric.I may electrify a line that has steam locos so they can be replaced.None of the trains I built myself are grouped.
Where formerly competing lines were acquired there are certainly opportunities to consolidate.But I'm a bit hazy on how to move things from individual lines to shared infra.
Doesn't serving intermediate stations with a passenger train lead to the passengers getting off the first chance they get and lowering your income?

[Simons Mith]

'Doesn't serving intermediate stations with a passenger train lead to the passengers getting off the first chance they get and lowering your income?'

Ye-es, but new ones get on. As a transport service, you should not care how far your passengers travel. 100 passengers each travelling 4 tiles is roughly as good as 4 passengers travelling 100 tiles. That's how you can make a profitable bus service inside a single city; as long as the buses spend all their time at least 50% full, and new passengers are constantly getting on and off, that's sufficient. I often have bus stops every 6 tiles or so with two sets of buses running clockwise and anti-clockwise routes. Similarly, as long as your trains are running full, you do not need to care what the station ratings are. Full trains are the more important factor, and higher stations ratings are merely a means to that end.

For long distance shipments, a good rule of thumb is to make the shipping distance 4x your vehicle's speed in mph, for 'perishable' goods like mail, passengers, food, and 7x for 'non-perishables' like coal, oil, wood etc. If source and distination are visible on-screen at the same time, use trucks to serve them - if you bother at all - not trains. Trains work best for transport distances of several screens.

But, the reason I posted was to comment on your Gondborough game. It's on the edge of saveable - maybe - but it's a click-fest, which isn't really in the spirit of OTTD! You have to demolish at least 100 stations and scrap all the unused track you possibly can, and send almost every empty train to depot to reduce running costs, and sell as many as possible of the old trains - except for a few that are wildly profitable. Normally I permit building while paused. In your game - assuming this is your saved settings and not a result of my recent mucking about with my own config settings, the only way I could find time to do all that was needed was to F1-unpause, click-demolish, F1-re-pause as quickly as I could. It works, barely, but it's not really in the spirit of the game. And if you take more than a couple of months game time to finish spring cleaning, you drop several millions in dept and won't recover. You've got to get those infrastructure maintenance costs down so fast that trying to achieve it is comical. I used the station view to sort the stations by lowest rating, which made it quicker to find the many dozens of non-functioning unconnected stations the AI has bequeathed you (perversely, they're all at the TOP of the list, because having never seen an ounce of cargo they've never been assigned a rating.). Similarly I sorted the trains by age so I could prioritise them.

Finally, on many of the most profitable lines, you have excess transport capacity. You can often afford to send as many as a third of the trains to depot, which means you're not paying their running costs, and then the remainder can run closer to full capacity without having multiple trains waiting at stations to fill up. On lines where there's only one train, I left them running regardless of reliability score, but wherever there were two or more trains I'd generally get rid of all but one or two because they just hold one another up. One train with 0% reliability may be slow, but as long as it's not holding anything else up, and as long as it's making a profit, there's more important things to worry about on this network.

Interesting challenge, though, I will say. Thanks.

[Baldy's Boss]

In the Rartown game,I just succumbed to temptation and bought my closest competitor for almost nothing plus assumption of debt I paid off.Their trains are numbers 15-26 of my new list.
I note that this had the same algorithm as those diesel lines in the Gondborough game with all the junctions...with over a thousand passengers on the platform at Flartbourne,I went around and lengthened the platforms and switched each linked train to a faster loco with an additional passenger car.Would cloning the trains for additional service also work?

[Baldy's Boss]

'Doesn't serving intermediate stations with a passenger train lead to the passengers getting off the first chance they get and lowering your income?'

Ye-es, but new ones get on. As a transport service, you should not care how far your passengers travel. 100 passengers each travelling 4 tiles is roughly as good as 4 passengers travelling 100 tiles. That's how you can make a profitable bus service inside a single city; as long as the buses spend all their time at least 50% full, and new passengers are constantly getting on and off, that's sufficient. I often have bus stops every 6 tiles or so with two sets of buses running clockwise and anti-clockwise routes. Similarly, as long as your trains are running full, you do not need to care what the station ratings are. Full trains are the more important factor, and higher stations ratings are merely a means to that end.

For long distance shipments, a good rule of thumb is to make the shipping distance 4x your vehicle's speed in mph, for 'perishable' goods like mail, passengers, food, and 7x for 'non-perishables' like coal, oil, wood etc. If source and distination are visible on-screen at the same time, use trucks to serve them - if you bother at all - not trains. Trains work best for transport distances of several screens.

But, the reason I posted was to comment on your Gondborough game. It's on the edge of saveable - maybe - but it's a click-fest, which isn't really in the spirit of OTTD! You have to demolish at least 100 stations and scrap all the unused track you possibly can, and send almost every empty train to depot to reduce running costs, and sell as many as possible of the old trains - except for a few that are wildly profitable. Normally I permit building while paused. In your game - assuming this is your saved settings and not a result of my recent mucking about with my own config settings, the only way I could find time to do all that was needed was to F1-unpause, click-demolish, F1-re-pause as quickly as I could. It works, barely, but it's not really in the spirit of the game. And if you take more than a couple of months game time to finish spring cleaning, you drop several millions in dept and won't recover. You've got to get those infrastructure maintenance costs down so fast that trying to achieve it is comical. I used the station view to sort the stations by lowest rating, which made it quicker to find the many dozens of non-functioning unconnected stations the AI has bequeathed you (perversely, they're all at the TOP of the list, because having never seen an ounce of cargo they've never been assigned a rating.). Similarly I sorted the trains by age so I could prioritise them.

Finally, on many of the most profitable lines, you have excess transport capacity. You can often afford to send as many as a third of the trains to depot, which means you're not paying their running costs, and then the remainder can run closer to full capacity without having multiple trains waiting at stations to fill up. On lines where there's only one train, I left them running regardless of reliability score, but wherever there were two or more trains I'd generally get rid of all but one or two because they just hold one another up. One train with 0% reliability may be slow, but as long as it's not holding anything else up, and as long as it's making a profit, there's more important things to worry about on this network.

Interesting challenge, though, I will say. Thanks.

Which save of Gondborough did you work from?...June 14th,July 31st,or August 26th?(The earlier I fork the more rescue work I have to do over).

A lot of the stranded stations of course were created by my acquiring a business before it had a chance to complete its construction plan.

If 100 passengers are replaced by 4 that go the rest of the way to the next station with 100,you're out a lot of passenger-miles.
I've never permitted building while paused.

[Baldy's Boss]

The latest save of the Plonnville game.The Charhill Woods station came with an acquisition,it was by a productive forest I wanted to serve,so I used it.It turns out it has those one-way signals you Deluxe initiates love,and I couldn't do my standard dedicated-track pair...I had to add a turning loop to the station I built at the sawmill.The third track at the sawmill is for connecting the Harnington forest,though I also want room to ship the goods over to Fondham or my HQ city of Plonnville.

In this game the company I'd be thinking of buying at this point is the quickly-growing but still cheap Fast Transport.Anyone want to advise on whether such a merger would be worth it?

Also below,I started a game with a much smaller world and much earlier start date than usual...at the mercy of the UKRS and NARS trainsets at the moment.I wonder what conventional wisdoms are less applicable in these circumstances.

[Transportman]

'Doesn't serving intermediate stations with a passenger train lead to the passengers getting off the first chance they get and lowering your income?'

It is not a simple yes/no question, and the answer might even change as the game progresses, as there are two effects with opposite effects. The first effect is a reduced income because of the longer travel time (when playing with CargoDist in recent nightlies/testing releases) or the shorter distance (both with and without CargoDist). The second effect is an increase of income as there is more cargo to transport on your network. For intermediate stations at small towns the first effect will dominate, but in the long run that small town can grow, encapsulating the station and then the second effect can dominate.

[Baldy's Boss]

Again,thanks for the advice so far.I'm not sure just where to fork Gondborough,but I'd been playing one-handed (didn't know the F1 trick) and the demolitions of extra stations should go quicker.On inspection I realize that the Buntburg I was advised to close before was an acquired,stranded station I hadn't noticed,and not the goods-receiving station that's given me such headaches (Buntburg West).

In the Old Plonnville game I've now bought 75% of the purple-colored second-place company (Fast Transport or Speedy Transport but I don't know what it will be called on next load),I'm open to advice on whether to buy out the rest.Two of its trains have been stranded in depots since they were built because of right-angle-turn geometry,I can fix that on a takeover.The company seems to be growing fast and with good operating margins.Meanwhile I have steel trains now lying in wait to collect the product of that competitor's iron-ore trains.

In the Rartown game...well,I actually broke down and bought some buses.I took a look at Mardston,the world's current largest city,and saw an intermodal station with a banner that said ChooChoo was building tracks and pathfinding...then I realized it was in my own colors.
I bought the company that was building it,and had also built a bus station in Mardston (apparently taking the form of bus shelters on the sidewalks of several blocks).The intermodal station included a road depot but was not linked to Mardston's streets.I put in the road link,and another to nearby Trendwood,where I built a bus station.The five buses now waiting in Mardston Woods depot have orders to Trendwood,Mardston,Mardston Woods (not sure if that's the best order).
The AI also built one of those straight,signaled track pairs between two four-way junctions that I've remarked on before.This runs between Mardston and not-too-far Menston,the world's second largest city,which I want to link,though it's roughly perpendicular to a path between them.I want to plug the track from Mardston Woods into one end of that straight and tracks from a Menston station into the other end...are there any particular tricks to getting the right interface between tracks and signals?
Other nearby cities (I hope people wanting to go to Kindingville or Gindingville have checked their tickets carefully) can also be linked in.

And I still need advice on the Buborough & Penbourne line,uploaded previously.Why is neither train near Buborough station prepared to proceed straight down its track?Is there anything keeping those trains and the two down near Penbourne station from proceeding through unoccupied signal blocks and switching to the other track as needed to avoid oncoming traffic?

[Baldy's Boss]

I went back to the June 14th Gondborough save (the first one uploaded) and played forward to August 10th (here uploaded).I found 83 stations not connected to others,which I demolished.Trains 3 and 164-et-al have gotten their electrification fixed,and I sold 13-17 as in the previous effort.the hopper cars from 13 and 16 are in the depot at Finnpool Woods and I electrified part of that line (it all has to be done for new locos).I replaced two old steamers,one on a previously electrified line and one on a short line I just electrified,and again ordered trains 18,117,118,and 160 to depots along with Group 27 and some others.
Doom still seems certain at the end of August.

[Eddi]

i see only one way out of this mess: remove every 2nd track, and use passing loops instead. should massively cut back both on your track and your signal costs

the ideal flow for equidistant passing loops is every second loop should be empty, and the other ones should have two trains in it. at each (2-track) loading station a train should enter exactly when the other one is trying to leave from full load, and the delivering station should be one track at passing loop distance. only use double track if the distance of passing loops gets shorter than one train length.

a system with 2 trains only needs the loading station double tracked, otherwise single tracks
a system with 3 trains needs loading station double tracked and a double track section at 1/2 of the way to the unloading station
a system with 4 trains needs loading station double tracked, a double track section at 1/3 of the way and a double track section at 2/3 of the way.
etc.

so for n trains you need (n-2) loops at 1/(n-1), 2/(n-1), ..., (n-2)/(n-1)

[Baldy's Boss]

i see only one way out of this mess: remove every 2nd track, and use passing loops instead. should massively cut back both on your track and your signal costs

the ideal flow for equidistant passing loops is every second loop should be empty, and the other ones should have two trains in it. at each (2-track) loading station a train should enter exactly when the other one is trying to leave from full load, and the delivering station should be one track at passing loop distance. only use double track if the distance of passing loops gets shorter than one train length.

a system with 2 trains only needs the loading station double tracked, otherwise single tracks
a system with 3 trains needs loading station double tracked and a double track section at 1/2 of the way to the unloading station
a system with 4 trains needs loading station double tracked, a double track section at 1/3 of the way and a double track section at 2/3 of the way.
etc.

so for n trains you need (n-2) loops at 1/(n-1), 2/(n-1), ..., (n-2)/(n-1)

So what selection from the signal menu goes with this?
The AIs tend to build much shorter signal blocks than I do,but I am no signalling expert (again,i don't know how to make the trains in the Buborough game go straight forward!)

[Eddi]

So what selection from the signal menu goes with this?

rule of thumb: use one-way path signals (rightmost entry) at the sidings, and two-way path signals at the stations.

Code: Select all

    /<------\
---/------->-\---

("<" indicates the location and direction of the signal)

[Simons Mith]

My 'rescued' Gondborough project took 1+1/4 game years before I was officially prepared to declare the company saved, although I could tell the rescue was working after about 9 months. I did start from the August save, which was in the direst of dire straits. I ditched all but one of the passenger services, I think and hence ripped up almost all that costly electrified passenger track. And by the time I was done Gondborough transport wasn't so much 'lean' as emaciated. 113 stations remaining, 109 trains. Some of the steamers have literally taken more than a year to ship their final cargos and get back to their depots. In hindsight I should have just ordered them straight to the nearest depot. Allowing them to complete their orders was more of a hindrance than a help. Some of them would still have taken months to get to a depot. Getting the company back on its feet - well, if you aren't clearly on the way to recovery inside a year, you've probably been too slow to save it at all. I do kinda feel that having to unpause-click-pause again in order to have time to do all that is needed is such a breach of the spirit of the rules anyway you might as well just permit building while paused.

Out of my initial 2.5 million, (before the company disappeared into the red) I bought about eight fast, reliable new locos to replace the old clunkers on the most profitable lines (1 train replacing 3-5). I cloned the clunkers' orders, but did not send the replacements out until the clunkers had limped to a depot. I had to buy these new trains straight away because I knew if I left it I'd have no cash to buy them. I did get a bit carried away and scrapped a couple of profitable lines I could have saved, but when you've got to literally halve your infrastructure costs in a year, erring on the side of scrapping too much was far preferable.

The buses were abandoned to their own devices. Almost all two-platform stations were reduced to one. All trains routes of less than 100 tiles were scrapped. After all, when you're paying 1000 per tile per year maintenance here, all trains on a route of that length need to be making at least /200/ grand a year, or they're part of the problem, not part of the solution! And finally, almost every train less than five tiles long, (bar about two) and some that were longer, were also scrapped.

As of Nov 2004 profits are now greater than costs again, and I could see about electrifying and replacing all the diesels. Probably fixing all those out-of-date buses, and the last few ageing trains is the next priority. Even now the company is still dipping into the red every month. As soon as that's done, electrifying is the priority, and that has to be done ASAP before the remaining diesels, even the new ones, become unreliable.

As I already said, 'interesting'.

Gondborough Rail Group, 19th Nov 2004.png (29.56 KiB) Viewed 2841 times

[Eddi]

btw, electrifying actually increases the problem as the non-linear part of the cost is per railtype. so having 50% electrified is optimal in that respect (assuming there is no difference in base cost between a single rail tile and a single electrified tile)

[Baldy's Boss]

My 'rescued' Gondborough project took 1+1/4 game years before I was officially prepared to declare the company saved, although I could tell the rescue was working after about 9 months. I did start from the August save, which was in the direst of dire straits. I ditched all but one of the passenger services, I think and hence ripped up almost all that costly electrified passenger track. And by the time I was done Gondborough transport wasn't so much 'lean' as emaciated. 113 stations remaining, 109 trains. Some of the steamers have literally taken more than a year to ship their final cargos and get back to their depots. In hindsight I should have just ordered them straight to the nearest depot. Allowing them to complete their orders was more of a hindrance than a help. Some of them would still have taken months to get to a depot. Getting the company back on its feet - well, if you aren't clearly on the way to recovery inside a year, you've probably been too slow to save it at all. I do kinda feel that having to unpause-click-pause again in order to have time to do all that is needed is such a breach of the spirit of the rules anyway you might as well just permit building while paused.

Out of my initial 2.5 million, (before the company disappeared into the red) I bought about eight fast, reliable new locos to replace the old clunkers on the most profitable lines (1 train replacing 3-5). I cloned the clunkers' orders, but did not send the replacements out until the clunkers had limped to a depot. I had to buy these new trains straight away because I knew if I left it I'd have no cash to buy them. I did get a bit carried away and scrapped a couple of profitable lines I could have saved, but when you've got to literally halve your infrastructure costs in a year, erring on the side of scrapping too much was far preferable.

The buses were abandoned to their own devices. Almost all two-platform stations were reduced to one. All trains routes of less than 100 tiles were scrapped. After all, when you're paying 1000 per tile per year maintenance here, all trains on a route of that length need to be making at least /200/ grand a year, or they're part of the problem, not part of the solution! And finally, almost every train less than five tiles long, (bar about two) and some that were longer, were also scrapped.

As of Nov 2004 profits are now greater than costs again, and I could see about electrifying and replacing all the diesels. Probably fixing all those out-of-date buses, and the last few ageing trains is the next priority. Even now the company is still dipping into the red every month. As soon as that's done, electrifying is the priority, and that has to be done ASAP before the remaining diesels, even the new ones, become unreliable.

As I already said, 'interesting'.

Gondborough Rail Group, 19th Nov 2004.png

At the current game year getting a faster,reliable locomotive generally requires either electrifying a track (BR/SH "87" to replace a steam or diesel engine) or a longer train (the double-headed IC125 is the only non-electric fast and still reliable,and the TGV/Eurostar are the speed upgrade for slower electrics)...with the AI-built stations typically allowing no room for trains to get longer.The Drontbourne (trains 253/254,though I still have to get the northbound to switch off the main track earlier so it doesn't kep the southbound waiting in Drontbourne station),Ladhattan (Train 9,though I have to get a double-header and room for it now),and Tondhead Cross (Group 11) passenger services were all doing very well that I could see.What would you say are the best freight routes?

[Baldy's Boss]

I was able to upgrade Train 9 to a Eurostar and lengthen both the platforms for it,while taking a turn out of its route.As of save below I've made a "how the hell didn't I notice that" route adjustment that enables taking a turn and tunnel out of the route of Train 11,which ALSO made over 600K last year,and extending the line of the new track through past the depot would take out three more turns.However,while lengthening the Sondinghead platform is trivial,the road behind Ladhattan station is on a slope,and Train 11's current platform there can't be extended back into it.I haven't checked the newly vacated central track for extendability.So sending 11 to the depot for a Eurostar still has issues.Meanwhile,Ladhattan's third passenger train,Train 10,has always been a dog,currently broken down on a deadhead run with no passengers back to where there are over 340 waiting.That one I can see ripping out in its entirety.

[audigex]

Hmm, interesting new gametype, anyone?

You make a scenario of a company in trouble and upload it to BaNaNaS, people have to try to turn it around in the shortest space possible/with the minimum amount of reduction of services etc. Could be interesting

Either way, this thread needs pics, I can't use OpenTTD at work

[planetmaker]

Hmm, interesting new gametype, anyone?

You make a scenario of a company in trouble and upload it to BaNaNaS, people have to try to turn it around in the shortest space possible/with the minimum amount of reduction of services etc. Could be interesting

Either way, this thread needs pics, I can't use OpenTTD at work

Sounds like a typical use for a scenario

If one would make a special-taylored, tiny game script to check for the victory condition (last year income, none of the crucial settings changed) it could be quite perfect

[audigex]

Perhaps gametype was the wrong word - metagame, maybe?

[Baldy's Boss]

i see only one way out of this mess: remove every 2nd track, and use passing loops instead. should massively cut back both on your track and your signal costs

the ideal flow for equidistant passing loops is every second loop should be empty, and the other ones should have two trains in it. at each (2-track) loading station a train should enter exactly when the other one is trying to leave from full load, and the delivering station should be one track at passing loop distance. only use double track if the distance of passing loops gets shorter than one train length.

a system with 2 trains only needs the loading station double tracked, otherwise single tracks
a system with 3 trains needs loading station double tracked and a double track section at 1/2 of the way to the unloading station
a system with 4 trains needs loading station double tracked, a double track section at 1/3 of the way and a double track section at 2/3 of the way.
etc.

so for n trains you need (n-2) loops at 1/(n-1), 2/(n-1), ..., (n-2)/(n-1)

I thought I posted this already.

In the Plonnville game,I've tried setting up a couple of signalled dual-track sections between single-track platforms on the goods run to Plonnville (trains 78,80,and 81 serve it at present...not sure how to tweak it for most efficiency).However,one of the trains recently entered the signal block containing the tunnel by the farm despite another train being pulled up to the signal on the far side of the tunnel,rather than taking the parralel track.Is the problem the type of signal,or the shape of the signal block,or both?